id
stringlengths
20
52
title
stringlengths
3
459
abstract
stringlengths
0
12.3k
classification_labels
sequence
numerical_classification_labels
sequence
http://arxiv.org/abs/2105.05542v1
!Qué maravilla! Multimodal Sarcasm Detection in Spanish: a Dataset and a Baseline
We construct the first ever multimodal sarcasm dataset for Spanish. The audiovisual dataset consists of sarcasm annotated text that is aligned with video and audio. The dataset represents two varieties of Spanish, a Latin American variety and a Peninsular Spanish variety, which ensures a wider dialectal coverage for this global language. We present several models for sarcasm detection that will serve as baselines in the future research. Our results show that results with text only (89%) are worse than when combining text with audio (91.9%). Finally, the best results are obtained when combining all the modalities: text, audio and video (93.1%).
[ "Visual Data in NLP", "Speech & Audio in NLP", "Sentiment Analysis", "Stylistic Analysis", "Multimodality" ]
[ 20, 70, 78, 67, 74 ]
SCOPUS_ID:84874514594
" Dude, it's not a appropriate word" : Negotiating word meanings, language ideologies, and identities in a literature discussion group
This study explored how ideologies of language and literacy and social and academic identities were constructed and contested during a literature discussion. In the event, a group of five (and later six) boys in a fourth grade bilingual classroom attempt to identify an unknown word in their novel: booger. Microethnographic discourse analysis and analyses of participants' movements across the classroom were conducted, and interpretations were informed by spatializing theories of social practice and identities. Findings illustrate how the students' and teacher's physical and discursive moves involved the negotiation of multiple ideologies of language learning. Findings suggest the need for a deeper understanding of the micro-level processes in which academic and social identities and learning opportunities for students are forged. © 2013 Elsevier Inc.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:84864055289
" Speak English!" A prescription or choice of English as a lingua franca in Ghanaian schools
One of the common practices in many basic schools in Ghana is the constant reminder to students to speak English at all times, and the threat of sanctions to those who do not abide by this language regulation. Considering that Ghana is a multilingual country, one would have thought that any of the Ghanaian languages can be used by students at different times. Using self-reported and interview data, the paper sets out to find out whether Ghanaian students speak English as a choice of lingua franca or due to some ideological underpinnings. The discussion opens a window into the motivations that move Ghanaian students to speak English both in and out of the classroom. The paper has implications for education, language policy and the National Literacy Acceleration Programme, NALAP. © 2012 Elsevier Inc.
[ "Multilinguality" ]
[ 0 ]
SCOPUS_ID:84875601085
" Why am I paraphrasing?": Undergraduate ESL writers' engagement with source-based academic writing and reading
One of the most common and vital areas of coverage in second language (L2) writing instruction is writing from sources, that is, the process of reading source text material and transferring content from that reading to writing. Research as well as everyday practice in the classroom has long shown that working with source texts is one of the most challenging of all academic literacy activities for L2 writers. This is particularly true in the domain of paraphrasing, an important and yet complicated device for the treatment of source text material. While the procedures involved in paraphrasing source text material may appear simple, the enactment of those procedures is a complex and often elusive experience for L2 writers. In this article we discuss a study of two mainland Chinese students' engagement with paraphrasing in an undergraduate academic writing course, with a particular focus on their understanding of the purposes and functions of paraphrasing and how such understanding influenced their paraphrasing practices. Our results reveal a multilayered relationship between the students and paraphrasing and contribute to the paraphrasing literature by drawing greater attention to paraphrasing from students' perspectives. © 2013 Elsevier Ltd.
[ "Paraphrasing", "Text Generation" ]
[ 32, 47 ]
SCOPUS_ID:84936757377
" there are no words " and " my language is full of words" continuous and joints in the theory of language christophe tarkos
[ "Linguistics & Cognitive NLP", "Linguistic Theories" ]
[ 48, 57 ]
SCOPUS_ID:85031286739
"'Tis my muse will have it so": Four Dimensions of scatology in Molloy
Writers have long employed scatology for shock value, satire, grotesque humor, and as a stimulus for philosophizing. In Molloy, Samuel Beckett employs scatology not only for satire but as a touchstone in examining the themes of language, creativity, religion, and existentialism. Beckett uses excrement as a metaphor to pursue his theories of language as excess. This leads naturally to the topic ofcreativity, which psychologists had associated with defecation in the infant mind, and to mythology, where we also find the association with creation. In looking at religion, the scatological can be elevated alchemically through psychology and mythology, or it can be used satirically to bring religion down-Molloy traverses both paths, scrutinizing Catholicism in particular. Finally, excrement functions in the novel as a leveler, both among humans and of humans, suggesting that much of the sentiment behind the shame, disgust, and evasion of excrement is rooted in existential anxiety.
[ "Linguistics & Cognitive NLP", "Linguistic Theories" ]
[ 48, 57 ]
SCOPUS_ID:85062969294
"A Cruel King" Is Not the Same as "A King Who Is Cruel": Modifier Position Affects How Words Are Encoded and Retrieved From Memory
We examined whether the position of modifiers in English influences how words are encoded and subsequently retrieved from memory. Compared with premodifiers, postmodifiers might confer more perceptual significance to the associated head nouns, are more consistent with the "given-before-new" information structure, and might also be easier to integrate because the head noun is available before the modifications are encountered. In 4 experiments, we investigated whether premodified (the cruel and merciless king), and postmodified (the king who was cruel and merciless) noun phrases (henceforth, NPs) could induce variations in ease of subsequent retrieval. In Experiments 1, 2, and 3, participants used more pronouns (he), as opposed to full descriptions (the king) to refer to postmodified NPs than to unmodified competitors, but pronominal reference to premodified NPs and unmodified competitors did not differ, suggesting that postmodified NPs are more accessible in memory. When the data from all 3 experiments were combined, we also observed significantly more pronominal reference to post- than to premodified NPs, as well as a greater increase in pronominal reference rates between postmodified NPs and unmodified competitors than between premodified NPs and unmodified competitors. In Experiment 4, words following critical pronouns were read faster when the pronouns referred to modified than to unmodified NPs, and also when the pronouns referred to post- rather than premodified NPs. Taken together, our results show enhanced retrieval facilitation for postmodified NPs compared with premodified NPs. These results are the first to demonstrate that the linear position of modifications results in measurable processing cost at a subsequent point. The results have important implications for memorybased theories of language processing, and also for theories assigning a central role for discourse status and information structure during sentence processing.
[ "Linguistics & Cognitive NLP", "Information Retrieval", "Linguistic Theories" ]
[ 48, 24, 57 ]
SCOPUS_ID:85142329397
"A Large Playground": Examining the Current State and Implications of Conversational Agent Adoption in Organizations
This study explores the short- and long-term expectations about adoption of conversational agents in the organizational frontline. Drawing from in-depth interviews with managers and developers in organizations that have implemented these agents, it sheds light on how the deployment of and collaboration with technology-based autonomous agents influences service activities, and expectations of changes in organizational processes. The interviews revealed that implementations are done on a rather iterative and experimental basis, often balancing the current limits of technology while aiming at improved efficiency, augmenting the work of frontline employees, and meeting the ever-growing demands from consumers.
[ "Natural Language Interfaces", "Dialogue Systems & Conversational Agents" ]
[ 11, 38 ]
http://arxiv.org/abs/2112.13800v1
"A Passage to India": Pre-trained Word Embeddings for Indian Languages
Dense word vectors or 'word embeddings' which encode semantic properties of words, have now become integral to NLP tasks like Machine Translation (MT), Question Answering (QA), Word Sense Disambiguation (WSD), and Information Retrieval (IR). In this paper, we use various existing approaches to create multiple word embeddings for 14 Indian languages. We place these embeddings for all these languages, viz., Assamese, Bengali, Gujarati, Hindi, Kannada, Konkani, Malayalam, Marathi, Nepali, Odiya, Punjabi, Sanskrit, Tamil, and Telugu in a single repository. Relatively newer approaches that emphasize catering to context (BERT, ELMo, etc.) have shown significant improvements, but require a large amount of resources to generate usable models. We release pre-trained embeddings generated using both contextual and non-contextual approaches. We also use MUSE and XLM to train cross-lingual embeddings for all pairs of the aforementioned languages. To show the efficacy of our embeddings, we evaluate our embedding models on XPOS, UPOS and NER tasks for all these languages. We release a total of 436 models using 8 different approaches. We hope they are useful for the resource-constrained Indian language NLP. The title of this paper refers to the famous novel 'A Passage to India' by E.M. Forster, published initially in 1924.
[ "Language Models", "Semantic Text Processing", "Representation Learning" ]
[ 52, 72, 12 ]
SCOPUS_ID:45449100949
"A good Arab is not a dead Arab - A racist incitement": On the accessibility of negated concepts
According to the teachings of negation in psycholinguistics, concepts within the scope of negation are eradicated from the mental representation and replaced by available antonyms. Thus, given enough processing time, He is not alive is represented as He is dead. However, a systematic look at natural language use suggests that this is not always the case. Instead, concepts within the scope of negation often remain accessible to addressees so that they can refer to them later on in the discourse ("Not alive but evolving", 1 "It's not alive but it was [alive]", 2 "Not dead but dying" 3 ). That concepts within the scope of negation are not suppressed unconditionally is established here by both offline and online experimental data showing that, when relevant either to early or to late context, concepts within the scope of negation are not discarded. Instead they remain active in the minds of speakers and listeners who integrate them into their discourse representation.
[ "Linguistics & Cognitive NLP", "Ethical NLP", "Psycholinguistics", "Responsible & Trustworthy NLP" ]
[ 48, 17, 77, 4 ]
SCOPUS_ID:84904650745
"A little more ironic" - Voice quality and segmental reduction differences between sarcastic and neutral utterances
The presented production experiment analyzes the phonetic differences between neutral (i.e. sincere) and sarcastically ironic utterances in German. Results show in line with previous studies that sarcastic irony is expressed by longer utterance durations, lower and flatter F0 contours, and a lower intensity level. Moreover, extending previous findings, sarcastic irony is also characterized by a more variable (in tendency breathier) voice quality and a higher degree of segmental reduction, probably reflecting the speakers' dissociation from the wording of their utterances.
[ "Stylistic Analysis", "Multimodality", "Speech & Audio in NLP", "Sentiment Analysis" ]
[ 67, 74, 70, 78 ]
SCOPUS_ID:78149455858
"A new description of empire": Edmund Burke and the regicide republic of letters
In a departure from the extensive scholarship which focuses on Edmund Burke's writings on the British Empire, this essay explores Burke's diagnosis of an entirely new imperial formation: "the universal empire of the Regicide Republick of France." In order to account for the emergence of this novel form of empire and to explain the logic of its expansion, Burke returns to his early language theory from A Philosophical Enquiry (1757), modifying his analysis in the face of a burgeoning radical print culture. While the Enquiry celebrates the affective powers of language, Burke's Letters on a Regicide Peace (1795-97) deplores the transformative potential of print. Indeed, in Burke's final writings, "the Empire of Regicide" is inaugurated and sustained by a globalizing republic of letters. © 2010 Project MUSE®.
[ "Linguistics & Cognitive NLP", "Linguistic Theories" ]
[ 48, 57 ]
SCOPUS_ID:84884559741
"A simple gal here, longing for loving, mutual understanding and stable relationship": Gendered performances in online personal advertisements
This study examines the representations of femininity and masculinity in the discourse of 200 online personal advertisements by Malaysian men and women. Using quantitative and qualitative methodologies, this study investigates the ways in which identity is gendered in the personals - how male and female identities, via nominal group and clauses, may be associated with certain behavioural traits, emotions and social activities - and to demonstrate identity variations that are construed by these categorizations. Nominal group and clause (Halliday, 1994) form my points of entry and patterns are semantically classified according to attitude (Martin & White, 2005) and Social Actor categorization (Van Leeuwen, 2008). Advertisers appear to orient to what is expected in the context of courtship initiation in this particular genre, and are found to produce socially specific verbal practices that correspond to stereotypical constructions of gender identity. More significant, however, is the construal of an identity type that is specifically Malaysian - i.e. The simple person. © Universiti Putra Malaysia Press.
[ "Information Retrieval", "Text Classification", "Information Extraction & Text Mining" ]
[ 24, 36, 3 ]
SCOPUS_ID:84899845623
"A stinking filthy race of people inbred with criminality" A discourse analysis of prejudicial talk about Gypsies in discussion forums
Gypsies have been shown to be a group subject to extreme prejudice and discrimination in the UK. The current research explores how Gypsies are portrayed and talked about within UK discussion forums. A discourse analysis was conducted on three discussion forums concerning Gypsies and how they should be treated. The analysis identified the following strategies as being commonly used to express hatred towards, and to argue against right for, Gypsies: (1) referring to Gypsies as the 'other' who are abnormal; (2) constructing criminality as a key characteristic of Gypsies; (3) suggesting that some Gypsies are 'bogus', which was used to argue against all Gypsies; and (4) presenting Gypsies as outside of the law and given favouritism over settled British communities. The findings are discussed in light of existing literature surrounding the prejudice towards Gypsies and other minorities and suggestions for overcoming this prejudice are presented.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85131434932
"Abayomi, we are the revolution": Women's Rights and Samba at Rio de Janeiro
The advent of the feminist movement in the twentieth century made it possible for socially organized women to begin seeking for the recognition of their rights and the change of gender roles which were socially built. Women’s rights started to be recognized as a human right. However, criteria of race and class have always been relevant, and have provided privileged positions for white women in the pursuit and attainment of rights, while black women continue to be stigmatized, remaining in the base of the social pyramid. In this regard, this paper questions: What is the relation between feminism and the cultural manifestation of samba in Rio de Janeiro in the conquering of women’s emancipation these days? This study investigates the relationship between feminism and samba from the Discourse Analysis of the samba lyrics "Abayomi", composed by the women from the ÉPreta project, released in their album in 2017. It seeks to identify the knowledge about human rights produced from the emancipatory cultural processes led by black women samba singers, who use the samba sung and composed by them as a tool for emancipation in human rights and for democracy today. Monica Graciela Zoppi-Fontana’s work was used as a theoretical reference. In her studies, the author analyzes the discourse from the categories of the French matrix Discourse Analysis, but she adds the categories of gender and race, especially when discussing the place from where these women speak and how these bodies’ markers, gender and race, provide a determining context in their discourses
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85099478843
"Adolescents are Terrorizing the North of the City" The emergence of street gangs as a media phenomenon in Quebec (1987-1989)
Before the development of the scientific field that deals with the institutional status of street gangs in Quebec, the category "street gang" was used in the media in a context of tension, struggle, and profound lability. Our objective was to take a genealogical perspective in analyzing both the struggles and instabilities that led to the emergence of this concept and the discursive mechanisms by which such groups managed to challenge public authorities and create the view that they are stable and necessary. By going beyond the moral panic that was part of the context in which street gangs began to be discussed in the media, looking at the "discursive event" makes it possible to analyze the discourse that led to the concept of the street gang becoming embodied in a concrete social existence. Our empirical work uses discourse analysis to examine articles from the newspapers La Presse and the Journal de Montreal published between 1987 and 1989.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85127891947
"Adverbs and functional heads" twenty years later: Cartographic methodology, verb raising and macro/micro-variation
Adverbs and Functional Heads: a Cross-Linguistic perspective (Cinque, Guglielmo. 1999. Adverbs and functional heads: A cross-linguistic perspective. New York & Oxford: Oxford University Press) - one of the founding works of "Syntactic Cartography"- combines some of the developments in Syntactic Theory from the 1980s and 1990s with insightful contributions from Linguistic Typology. This paper has two interrelated goals. First, it aims to review the fundamental theses of Cinque's monography of 1999 - which are far from controversial among scholars working in Cartography -; at the same time it provides conceptual support to them. Secondly, it aims to explore some methodological tools of Syntactic Cartography presented and discussed by Cinque, Guglielmo. 1999. Adverbs and functional heads: A cross-linguistic perspective. New York & Oxford: Oxford University Press, namely the so-called precedence-and-transitivity tests - after a brief discussion on methodology used to recognise the functional categories, namely the criterion by Jackendoff, Ray. 1972. Semantic interpretation in generative grammar. Cambridge, MA: MIT Press - and the use of the hierarchies as tools to detect intra and interlinguistic variation. With regard to this latter issue, the paper gathers data from Brazilian Portuguese, Canadian English and Colombian Spanish on verb raising. The discussion of the data not only favours Cinque, Guglielmo. 2017. On the status of functional categories (heads and phrases). Language and Linguistics 18(4). 521-576 recent updates of his theoretical approach to the cartography of the clause but also shows how Cartography offers a natural scenario for a methodological approach to both micro and macro-variation.
[ "Typology", "Syntactic Text Processing", "Multilinguality" ]
[ 45, 15, 0 ]
SCOPUS_ID:85019559408
"Alexa is my new BFF": Social roles, user satisfaction, and personification of the Amazon Echo
Amazon's Echo and its conversational agent Alexa open exciting opportunities for understanding how people perceive and interact with virtual agents. Drawing from user reviews of the Echo posted to Amazon.com, this case study explores the degree to which user reviews indicate personification of the device, sociability level of interactions, factors linked with personification, and influences on user satisfaction. Results indicate marked variance in how people refer to the device, with over half using the personified name Alexa but most referencing the device with object pronouns. Degree of device personification is linked with sociability of interactions: greater personification co-occurs with more social interactions with the Echo. Reviewers mentioning multiple member households are more likely to personify the device than reviewers mentioning living alone. Even after controlling for technical issues, personification predicts user satisfaction with the Echo.
[ "Natural Language Interfaces", "Dialogue Systems & Conversational Agents" ]
[ 11, 38 ]
SCOPUS_ID:85129748895
"Alexa, Do You Want to Build a Snowman?" Characterizing Playful Requests to Conversational Agents
Conversational Agents (CAs) such as Apple's Siri and Amazon's Alexa are well-suited for task-oriented interactions ("Call Jason"), but other interaction types are often beyond their capabilities. One notable example is playful requests: for example, people ask their CAs personal questions ("What's your favorite color?") or joke with them, sometimes at their expense ("Find Nemo"). Failing to recognize playfulness causes user dissatisfaction and abandonment, destroying the precious rapport with the CA. Today, playful CA behavior is achieved through manually curated replies to hard-coded questions. We take a step towards understanding and scaling playfulness by characterizing playful opportunities. To map the problem's landscape, we draw inspiration from humor theories and analyze real user data. We present a taxonomy of playful requests and explore its prevalence in real Alexa traffic. We hope to inspire new avenues towards more human-like CAs.
[ "Natural Language Interfaces", "Dialogue Systems & Conversational Agents" ]
[ 11, 38 ]
SCOPUS_ID:84879496567
"Aliança pelo Centro Histórico": Discursive strategies legitimating hygienization processes in the city of São Paulo
The aim of this paper is to investigate the discursive strategies used by the Aliança pelo Centro Histórico in the media in order to reveal the existing conflict between the rights of the disadvantaged groups and the financial interests that takes place in the central area of the city of São Paulo. The undertaken analysis is based on Critical Discourse Analysis (specially Fairclough and van Dijk). It reveals the existence of a process of dehumanization of homeless people, which legitimates both violence and a politics of expulsion of this group from the city center.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85115297834
"And i have voted for zou back then." online comments sections between text, discourse and community
Using examples from comment sections in the Swiss online news site watson.ch, we investigate the question whether such online reader communities show traces of communal relationships (in the sense of Max Weber) and if such an understanding is compatible with the concept of discourse communities. To this end, we first outline the conflicting theoretical assumptions that linguistic and sociological understandings of communities imply for the concept of discourse communities. Afterwards, we use selected online reader comments to show how traditional features of community formation can be detected empirically in the comment section. To conclude, we argue that Niklas Luhmann's systems theory can serve as a framework to integrate different types of community formation processes on an empirical and a theoretical level.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85040908545
"And this is the view from outside my window": On text and image interplay in university website blogs
The paper focuses on the institutional website as a complex genre with a relatively discontinuous inner structure, which is, however, coherent and cohesive, and unified by a common communication goal(s). The website is viewed as a discourse colony consisting of independent but related components realized in an array of subgenres, some of which are typical of the academic/institutional environment while others come from different discourse domains and are employed as embedded genres. The paper focuses on the blog as an embedded genre, its forms and functions within university websites, and particularly on its potentially multimodal character, i.e. the interplay of the verbal content of the blog and the non-verbal elements, esp. photographs, which co-create the producer's message to the addressee. Drawing upon the recently developed field of multimodal discourse analysis within Hallidayan Systemic Functional Linguistics, particularly Martinec and Salway's model, the paper explores the level to which the modes are integrated and the ways they contribute to meaningmaking in the genre.
[ "Visual Data in NLP", "Multimodality" ]
[ 20, 74 ]
SCOPUS_ID:84955306053
"Answer ka type kya he?" learning to classify questions in code-mixed language
Code-Mixing (CM) is defined as the embedding of linguistic units such as phrases, words, and morphemes of one language into an utterance of another language. CM is a natural phenomenon observed in many multilingual societies. It helps in speeding-up communication and allows wider variety of expression due to which it has become a popular mode of communication in social media forums like Facebook and Twitter. However, current Question Answering (QA) research and systems only support expressing a question in a single language which is an unrealistic and hard proposition especially for certain domains like health and technology. In this paper, we take the first step towards the development of a full-fledged QA system in CM language which is building a Question Classification (QC) system. The QC system analyzes the user question and infers the expected Answer Type (AType). The AType helps in locating and verifying the answer as it imposes certain type-specific constraints. We learn a basic Support Vector Machine (SVM) based QC system for English-Hindi CM questions. Due to the inherent complexities involved in processing CM language and also the unavailability of language processing resources such POS taggers, Chunkers, Parsers, we design our current system using only word-level resources such as language identification, transliteration and lexical translation. To reduce data sparsity and leverage resources available in a resourcerich language, in stead of extracting features directly from the original CM words, we translate them commonly into English and then perform featurization. We created an evaluation dataset for this task and our system achieves an accuracy of 63% and 45% in coarse-grained and fine-grained.
[ "Machine Translation", "Information Retrieval", "Information Extraction & Text Mining", "Question Answering", "Natural Language Interfaces", "Text Generation", "Text Classification", "Multilinguality" ]
[ 51, 24, 3, 27, 11, 47, 36, 0 ]
SCOPUS_ID:85039708455
"Apparently, women don't know how to operate doors": A corpus-based analysis of women stereotypes in the TV series 3<sup>rd</sup> Rock from the Sun
This paper explores how women stereotypes are discursively evaluated in the TV sitcom 3rd Rock from the Sun by paying attention to the societal, cultural and ideological values they convey. Following recent trends for the study of television series (Bednarek, 2010), the analysis is both qualitative and quantitative, adopting a Corpus- Assisted Discourse Analysis approach (Baker, 2006; Partington, 2004). The contextualised analysis of words that refer to women confirms that the sitcom writers of 3rd Rock from the Sun purposefully resort to stereotyping as a verbal strategy to create humour while conveying negative attitudes towards women.
[ "Discourse & Pragmatics", "Semantic Text Processing", "Ethical NLP", "Responsible & Trustworthy NLP" ]
[ 71, 72, 17, 4 ]
SCOPUS_ID:85139457201
"Are we still talking about the same thing?" MEG reveals perspective-taking in response to pragmatic violations, but not in anticipation
The current study investigates whether mentalizing, or taking the perspective of your interlocutor, plays an essential role throughout a conversation or whether it is mostly used in reaction to misunderstandings. This study is the first to use a brain-imaging method, MEG, to answer this question. In a first phase of the experiment, MEG participants interacted "live" with a confederate who set naming precedents for certain pictures. In a later phase, these precedents were sometimes broken by a speaker who named the same picture in a different way. This could be done by the same speaker, who set the precedent, or by a different speaker. Source analysis of MEG data showed that in the 800 ms before the naming, when the picture was already on the screen, episodic memory and language areas were activated, but no mentalizing areas, suggesting that the speaker's naming intentions were not anticipated by the listener on the basis of shared experiences. Mentalizing areas only became activated after the same speaker had broken a precedent, which we interpret as a reaction to the violation of conversational pragmatics.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
http://arxiv.org/abs/2208.08909v2
"Are you okay, honey?": Recognizing Emotions among Couples Managing Diabetes in Daily Life using Multimodal Real-World Smartwatch Data
Couples generally manage chronic diseases together and the management takes an emotional toll on both patients and their romantic partners. Consequently, recognizing the emotions of each partner in daily life could provide an insight into their emotional well-being in chronic disease management. Currently, the process of assessing each partner's emotions is manual, time-intensive, and costly. Despite the existence of works on emotion recognition among couples, none of these works have used data collected from couples' interactions in daily life. In this work, we collected 85 hours (1,021 5-minute samples) of real-world multimodal smartwatch sensor data (speech, heart rate, accelerometer, and gyroscope) and self-reported emotion data (n=612) from 26 partners (13 couples) managing diabetes mellitus type 2 in daily life. We extracted physiological, movement, acoustic, and linguistic features, and trained machine learning models (support vector machine and random forest) to recognize each partner's self-reported emotions (valence and arousal). Our results from the best models (balanced accuracies of 63.8% and 78.1% for arousal and valence respectively) are better than chance and our prior work that also used data from German-speaking, Swiss-based couples, albeit, in the lab. This work contributes toward building automated emotion recognition systems that would eventually enable partners to monitor their emotions in daily life and enable the delivery of interventions to improve their emotional well-being.
[ "Emotion Analysis", "Sentiment Analysis", "Multimodality" ]
[ 61, 78, 74 ]
SCOPUS_ID:85114084661
"Are you really voting for your favourites?" Multimodal genres and ideology in talent shows
This chapter presents an analysis of the multimodal discursive strategies employed in the creation of talent shows which have international editions and involve several contemporary "performance" spaces (TV, web, social networks) revealing a multifarious macro show genre which is not tied to a single media or web environment. A combined Systemic Functional Linguistics and multimodal corpus-assisted approach (Halliday, 1978, 1994; Baldry & Thibault, 2006; Kress, 2018) is used first to investigate the multimodal strategies employed in the 2018 season of the X Factor (UK) and, then, to analyse some of the language used by the official judges of the show. Considering talent shows in the framework of a more general discourse on evaluation, the paper explores how the construction of consensus and evaluation is built up via the use of the judges' language.
[ "Multimodality" ]
[ 74 ]
SCOPUS_ID:85071501648
"At home while you can." Discursive construction of home attachment in elderly people
The bond established between older people and their homes is a central concern in psychology. This study approaches to the phenomenon through discourse analysis on how place attachment is attributed to telecare users in Spain. The aim is to offer an approach to how place attachment is produced through talk during given interactions. The data come from a series of focus groups conducted in Catalonia (Spain). Three topics are derived from the analysis. The definition of home attachment as the desire to remain at home in old age is predominant. The analysis focuses in such a topic. The meanings associated to it is deepened, highlighting the discursive strategies with which two groups of actors produce it. Attention is drawn to both the interpretive repertoires integrating such strategies and their discursive functions.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85046621234
"Autopoet" project: A semantic anomalies generator or a new existence creator?
The article is devoted to the problems of computer generation of semantically abnormal texts possessing a grammatically connected structure, but entering into an obvious contradiction with the logic of objective reality. Using the example of the "Autopoet" service developed by Yandex showed that machine text generation is of interest not only in the applied aspect, but also as an instrument of aesthetic influence that allows rethinking the traditional paradigms of literary and art creation. Processing millions of search requests, "Autopoet" generates absurdist poems on their basis. Examination of the functioning principles of the "Autopoet" in the existential and phenomenological aspects allows concluding that, on the one hand, its poetic experiments embody a completely impersonal type of writing and, consequently, do not have a clear compositional and semantic structure. On the other hand, amorphous and rhizomatic computer "auto-text" every time as a building made of bricks, consists of self-sufficient microtexts, once entered by users in the search box and carrying a powerful existential charge, because each of them captures the unique moment of everyday human existence. Thus, the "Autopoet" alienating the verbal and existential acts from the subject who performs them, accumulates them in the space of the text, changing them into a pure, impersonal existence.
[ "Text Generation" ]
[ 47 ]
http://arxiv.org/abs/2104.08673v2
"Average" Approximates "First Principal Component"? An Empirical Analysis on Representations from Neural Language Models
Contextualized representations based on neural language models have furthered the state of the art in various NLP tasks. Despite its great success, the nature of such representations remains a mystery. In this paper, we present an empirical property of these representations -- "average" approximates "first principal component". Specifically, experiments show that the average of these representations shares almost the same direction as the first principal component of the matrix whose columns are these representations. We believe this explains why the average representation is always a simple yet strong baseline. Our further examinations show that this property also holds in more challenging scenarios, for example, when the representations are from a model right after its random initialization. Therefore, we conjecture that this property is intrinsic to the distribution of representations and not necessarily related to the input structure. We realize that these representations empirically follow a normal distribution for each dimension, and by assuming this is true, we demonstrate that the empirical property can be in fact derived mathematically.
[ "Language Models", "Semantic Text Processing", "Representation Learning" ]
[ 52, 72, 12 ]
SCOPUS_ID:85045849092
"Bacon wrapped cancer": The discursive construction of meat carcinogenicity
In 2015, the World Health Organization published a report on the carcinogenicity of red and processed meat (IARC, 2015. Carcinogenicity of consumption of red and processed meat. The Lancet Oncology 16(16). 1599-1600), attracting intense interest from both the general public and the scientific community. This study combines corpus approaches, Systemic Functional Linguistics and discourse analysis to investigate and compare scientific and animal rights movement reactions to the IARC 2015 report. Scientific reactions are exemplified by three research papers published immediately after the report; responses from animal rights campaigners are investigated through an analysis of texts taken from the website of the nongovernmental organization PETA. The aim is to explore how discourse not only describes, but also constructs meat carcinogenicity, in texts produced by two discourse communities (scientists and animal campaigners) which, for entirely different reasons, have an important stake in this issue. Qualitative (close reading) and quantitative (corpus-based) methods are combined, focusing on vocabulary, grammatical metaphor, and Appraisal (Martin, Jim and Peter White, 2005. The language of evaluation: Appraisal in English. Basingstoke: Palgrave Macmillan). The results show a high level of hybridity, discursive erasure (Stibbe, Arran, 2012. Animals erased: Discourse, ecology, and reconnection with the natural world. Middletown: Wesleyan University Press), and some substantial differences in the discourse reactions to the IARC report by the two sources, reflecting the ideologies and ethical assumptions they espouse in their approach to the announcement that red and processed meat can cause cancer.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:84940829377
"Between love and hatred": Pacifism, Gospel texts and images of Jesus in the writings of Y. T. Wu, 1918-1948
The idea of pacifism, which Y. T. Wu, a famous Christian Intellectual in Republican China, stuck to for at least 20 years, has been dealt with in the previous scholarship on his life and thought. The less-attended question concerning pacifism, however, is how Wu's understanding, interpretation and use of various Gospel texts in different periods influence his images of Jesus. This article, taking "reception history of the Bible" as its approach, investigates this question in light of Wu's life and the political contexts in which he lived from 1918 to 1948 in the following three dimensions: (1) Jesus' love and non-violence; (2) Jesus' hatred of sins as well as his view on the necessity of class straggle (jieji douzheng), revolution (geming) and even violence (wuli) in pursuit of justice; and (3) Jesus' hatred of sinners. Firstly, Wu from late 1910s until late 1930s firmly believed that the most significant image of Jesus was his love and non-violence reflected most obviously in the Sermon on the Mount recorded in the Gospel of Matthew. And even the deterioration of Sino-Japanese relation in the 1930s did not totally sway him from this perception. Secondly, expressed in the writings of Wu in the 1920s and 1930s, Jesus who loved also hated sins (rather than sinners yet); and even though Jesus preferred non-violence, he also advocated the so-called non-violent struggle and revolution for the sake of justice and love. There were some sayings of Jesus from the Gospels interpreted by Wu to justify the above assertions, such as "I have come to send fire on the earth" and "I have come to give...division". Since late 1930s, the periscope of Jesus' cleansing of the temple in the Gospels came to be used for demonstrating a violence that Jesus, under some circumstances, also exerted. And this image was able to co-exist with that of the non-violent Jesus in Wu's mind. Thirdly, from 1939 onward, Wu's image of Jesus who only hated sins gradually evolved into one who hated sinners as well. This was persuasively argued for in his debate with Margaret H. Brown, a Canadian missionary in China, in 1948. Nonetheless, even at that time Wu still insisted that Jesus only hated the "reality" of sinners but loved their "potentiality". On the whole, Wu's images of Jesus from 1918-1948 were constantly "between love and hatred".
[ "Visual Data in NLP", "Multimodality" ]
[ 20, 74 ]
SCOPUS_ID:84969752945
"Beyond GDP": Lessons from the discourses of institutional actors
New indicators "beyond GDP" have increasingly sparked interest among various actors, whose status, objectives and visions are very different. While the pioneer reflections on economic growth and GDP were first limited to environmentalist and activist movements, civil society and local policies, they have progressively entered national and international institutional spheres. The diversity of theoretical approaches and normative positions regarding the opportunity and motives of going "beyond GDP" makes it hard to clearly identify the stances of the actors and the power balances ruling the debates. We therefore ask: what does the current interest of institutional actors in "beyond GDP" issues mean? Are they a new rhetoric liable to elude a confrontation with the structural problems resulting from the crisis? Are they an opportunity window for relaunching societal debates that are hardly raised elsewhere? Or are they a real driver toward a paradigmatic change? We try to answer that question by analyzing the discourses of institutional actors (politicians, administration, technicians) involved or not in "beyond GDP" initiatives. We show that, at the institutional level, beyond GDP debates, while raising new societal issues, do not contribute to erode the central role of economic growth. Pragmatism dominates the debates, in that dominant interests are focused on short-term constraints and objectives where GDP growth remains pivotal. The involvement of actors in "Beyond GDP" debates reveals more a need to adapt public management and policies to new constraints rather than a critical reflection on the productivist model on which our economies have been built for more than sixty years.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85015192902
"Big data" and the electronic health record
OBJECTIVES: Implementation of Electronic Health Record (EHR) systems continues to expand. The massive number of patient encounters results in high amounts of stored data. Transforming clinical data into knowledge to improve patient care has been the goal of biomedical informatics professionals for many decades, and this work is now increasingly recognized outside our field. In reviewing the literature for the past three years, we focus on "big data" in the context of EHR systems and we report on some examples of how secondary use of data has been put into practice.
[ "Ethical NLP", "Responsible & Trustworthy NLP" ]
[ 17, 4 ]
SCOPUS_ID:84902077232
"Bin ich wohl etwas naiv gewesen.": Zur Rezeption empirischer Bildungsforschung in der Öffentlichkeit - Das Beispiel ELEMENT
The authors inquire into the rules and patterns of the public discourse on the results of empirical educational research as illustrated by the public response to the Berlin ELEMENT-Study. 37 press articles published during the first six months after publication of the study's results are analyzed on the basis of the knowledge-sociological discourse analysis. The press well reflects the many-facetted public discourse on the ELEMENT-Study. The results of this discourse analysis provide information on the dissemination to the public of research results and on their transformation in the sense of specific modes of trivialization. Thus, for instance, two conflicting interpretations of the political-public role of educational research are revealed: the concept of evidence-based policies vs. a division of tasks among science and politics. Furthermore, it can be shown that the discourse on the ELEMENT-Study orients itself mainly by the demands of the upper milieus. The results of the analysis provide impulses for a target-oriented input of research results into the public debate. Some of the initial conclusions are discussed.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:84991672189
"Bits", "Chunks" and "Channel-switching": Perceptions of Cantonese-English code-switching
This paper compares the code-switching (CS) practices and perceptions of three groups of Cantonese-English bilinguals with different social backgrounds: (1) university-educated Hong Kong locals, (2) second-generation migrants in English-speaking countries and (3) Hong Kong students who have spent a significant amount of time in English-speaking countries at school and/or university. The results show that far from agreeing on what constitutes a normal CS style, each of these groups has different normative perceptual contours, and distinct CS practices. This finding is shown to be significant within the social context of the Hong Kong speech community, which has reabsorbed a great deal of the Cantonese-English speaking diaspora in recent years. The consequences of this reabsorption are also shown to be important for our current models of CS variation.
[ "Code-Switching", "Multilinguality" ]
[ 7, 0 ]
SCOPUS_ID:85029128269
"Breakthrough into performance": How understanding communication as performance can transform teaching, learning, assessment and curriculum
The Engineering Communications Program (ECP) in the College of Engineering at Cornell University is gradually integrating a new understanding of communication. That understanding is technical and professional communication as performance. It is particularly responsive to many of the difficulties related to establishing an effective communications curriculum. First, it offers engineering faculty a common understanding of communication. Second, through focusing on those elements shared by all communicative performances - genre, context, identity, and purpose - it enables faculty to adapt their instructional approach in ways that acknowledge the diversity of possible performances while maintaining coherence. Third, it allows faculty to employ a useful methodology, discourse analysis, for doing assessment as educational research. This paper presents communication as performance, offers discourse analysis as a methodology for researching those performances, and suggests how communicative performance, when matched with discourse analysis, can provide a dynamic model for a new communications curriculum that emerges from within the community of practice that is engineering. © American Society for Engineering Education, 2006.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85065444333
"Brāhmana" as an honorific in "Indianized" mainland Southeast Asia: A linguistic approach
This article aims at demonstrating that the Old Khmer b/vrah originates from a syllabic depletion of the Sanskrit word brāhmana through a monosyllabization process, a widespread diachronic phenomenon among the Mon-Khmer languages of Mainland Southeast Asia. The paper will also show that this term must have been originally used as an honorific for deities and, consequently, for royalty. It therefore respectfully disagrees with two other current hypotheses according to which b/vrah would be an autochthonous Mon-Khmer word or would originate in the Sanskrit/Pali word vara- excellent, splendid, noble. After being borrowed from Sanskrit, the Old Khmer brah spread via a contact phenomenon: from Old Khmer to Old Siamese, from Old Siamese to Old Shan through the Thai Continuum, and from Old Shan to Old Burmese. The implications of this paper are twofold: firstly, it will sketch out a pattern for the historical relationships between different peoples of Mainland Southeast Asia; then, it will propose a first phase of Indianization in Southeast Asia, namely a local reconnotation of Indo-Aryan terms according to autochthonous socio-political contingencies, and consequently bring a draft answer to the Woltersian question: what is the local connotation of Indo-Aryan terms?
[ "Phonology", "Syntactic Text Processing" ]
[ 6, 15 ]
SCOPUS_ID:84902977568
"But I Am Not One to Judge Her Actions": Thematic and Discursive Approaches to University Students' Responses to Women Who Smoke While Pregnant
Qualitative methodologies offer various approaches to interpreting qualitative data. Here we consider how different approaches to interpreting the same data can be useful in learning about the scope and utility of qualitative methods and in exploring the role of reflexivity in analytic decision making and interpretation. We apply both thematic and discourse analyses to university students' responses to an open-ended question about women who smoke while pregnant. We show how our interpretations differ when analytic attention is paid to the content (thematic analysis) versus the rhetorical function (discourse analysis) of participants' responses. We also show how reflexivity, compatible with our discursive analysis, allowed us to identify the local discursive context in which the data were produced and therefore how participants oriented to this context. We use our learning experience as a way of showcasing the value of dynamic and reflexive approaches to qualitative data. © Taylor & Francis Group, LLC.
[ "Discourse & Pragmatics", "Explainability & Interpretability in NLP", "Semantic Text Processing", "Responsible & Trustworthy NLP" ]
[ 71, 81, 72, 4 ]
SCOPUS_ID:84865298004
"But Most of All, They Fought Together": Judicial Attributions for Sentences in Convicting Battered Women Who Kill
The present study provides a discourse analysis of judicial attributions about battered women in Canadian sentencing decisions involving women convicted of killing their abusive intimate partners. For cases in which the accused received a jail sentence, judges downgraded acts of previous partner violence by using minimizing descriptions and by emphasizing the mutuality of the violence and of substance abuse. These discourses mobilized doubt about the relationship as abusive and limited sympathy for the accused as a battered woman. Judges' descriptions formulated domestic abuse as discrete episodes of violence, attributed in many cases to alcohol rather than to an ongoing pattern of serious domestic abuse. These descriptions reinforced the accused's capabilities and strength, which served to diminish the opposing claim that she was trapped in a seriously abusive relationship. Recommendations include incorporating information about battered women's resistance efforts into traditional battered woman syndrome testimony and examining police decision making in cases of dual arrest. © The Author(s) 2012.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:62849092312
"But it's all true!": Commercialism and commitment in the discourse of organic food promotion
Debates over food politics provide insight into the convergence of commercial and political discourses. As the organic food market has grown, campaigners and independent producers have faced the dilemma of how far they should promote their cause using standard marketing language. We report on a research project which combined corpus analysis, interviews, and focus group discussions to investigate the discourse of organic food promotion in Britain, the thinking behind it, and how people react to it. We found growing convergence across the sector. Whether produced by supermarkets, small politically committed producers, or environmentalist campaign groups, the language used tends to be poetic, vague, dialogic, narrative, and emotive, with an emphasis upon bucolic imagery and consumer self-interest. Text producers assume that consumer attitudes can be easily manipulated by such an approach. Our focus group data however suggest both a critical resistance to marketing language in general, and that attitudes to food may be less amenable to manipulation through standard promotional techniques than is commonly assumed. Our findings contribute not only to an understanding of food politics and persuasive discourse more generally, but also to the development of discourse analytic methodology which integrates textual analysis with investigation of sender and receiver perceptions. © 2009 by Walter de Gruyter GmbH & Co. KG.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
http://arxiv.org/abs/2207.01483v1
"COVID-19 was a FIFA conspiracy #curropt": An Investigation into the Viral Spread of COVID-19 Misinformation
The outbreak of the infectious and fatal disease COVID-19 has revealed that pandemics assail public health in two waves: first, from the contagion itself and second, from plagues of suspicion and stigma. Now, we have in our hands and on our phones an outbreak of moral controversy. Modern dependency on social medias has not only facilitated access to the locations of vaccine clinics and testing sites but also-and more frequently-to the convoluted explanations of how "COVID-19 was a FIFA conspiracy"[1]. The MIT Media Lab finds that false news "diffuses significantly farther, faster, deeper, and more broadly than truth, in all categories of information, and by an order of magnitude"[2]. The question is, how does the spread of misinformation interact with a physical epidemic disease? In this paper, we estimate the extent to which misinformation has influenced the course of the COVID-19 pandemic using natural language processing models and provide a strategy to combat social media posts that are likely to cause widespread harm.
[ "Reasoning", "Fact & Claim Verification", "Ethical NLP", "Responsible & Trustworthy NLP" ]
[ 8, 46, 17, 4 ]
http://arxiv.org/abs/2004.12764v2
"Call me sexist, but...": Revisiting Sexism Detection Using Psychological Scales and Adversarial Samples
Research has focused on automated methods to effectively detect sexism online. Although overt sexism seems easy to spot, its subtle forms and manifold expressions are not. In this paper, we outline the different dimensions of sexism by grounding them in their implementation in psychological scales. From the scales, we derive a codebook for sexism in social media, which we use to annotate existing and novel datasets, surfacing their limitations in breadth and validity with respect to the construct of sexism. Next, we leverage the annotated datasets to generate adversarial examples, and test the reliability of sexism detection methods. Results indicate that current machine learning models pick up on a very narrow set of linguistic markers of sexism and do not generalize well to out-of-domain examples. Yet, including diverse data and adversarial examples at training time results in models that generalize better and that are more robust to artifacts of data collection. By providing a scale-based codebook and insights regarding the shortcomings of the state-of-the-art, we hope to contribute to the development of better and broader models for sexism detection, including reflections on theory-driven approaches to data collection.
[ "Ethical NLP", "Robustness in NLP", "Responsible & Trustworthy NLP" ]
[ 17, 58, 4 ]
SCOPUS_ID:84933502267
"Can i make a party, mum?" The development of requests from childhood to adolescence
This study presents how requests are acquired and developed over an eight-year period by an EFL learner in a foreign language setting, where target language pragmatics is not an issue dealt with in the classroom. In order to assess pragmatic development, a role-play requiring requests was used. This study has been triggered by the fact that longitudinal studies have commonly been considered very valuable, since development of the same participants can be traced over a long period of time. The development of requests has been followed by, first, examining what types of requests were produced by the learner at the different stages of pragmatic development; second, by analyzing the use of request modification; and, finally, by placing the learner's requests at different stages of development. The results seem to show that little development can be traced at very early stages of acquisition, and that it is not until Grade 11 that a development toward more pragmatically appropriate productions can be found.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:84947427295
"Can i very please borrow it?": Request development in young Norwegian EFL learners
With the introduction of the notion of communicative competence to second-language learning and teaching (Canale and Swain 1980), and the recognition of the role of pragmatic competence within it (Bachman 1990; Bachman and Palmer 1996), interlanguage pragmatics (ILP) research has gained in popularity. However, with a few notable exceptions (Achiba 2002; Barón Parés 2012; Ellis 1992; Rose 2000 and Rose 2009), ILP research has focused almost exclusively on adult learners, and even with that learner group, studies of pragmatic development have been comparatively rare (Kasper and Rose 2002). The present study set out to address a generally neglected area in ILP research: developmental patterns in speech acts-more specifically, the development of requests in young Norwegian EFL learners. The aims of the study were to identify specific request strategies that emerge at different stages of development and to explore learners' sensitivity to social power as a contextual factor. Three age groups of pupils (8, 10, and 12 years old) participated in this cross-sectional study. The data were collected through a short structured interview and role plays and analyzed in terms of the level of directness, the types of head acts, and their internal and external modification (Blum-Kulka et al. 1989). While the results revealed clear patterns of pragmalinguistic development with regard to the complexity of head acts and the use of alerters, supportive moves, and downgraders, little evidence of sociopragmatic development was found in the data. This exploratory study opens a number of avenues for further exploration of pragmatic development in young EFL learners.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85091654763
"Che cos'è? Va e de"?: Questions, answers and code-switching in two Swedish classes of Italian L2/LS
This conversation-analytic study investigates the use of code switching (henceforth CS) in three Italian L2 classrooms in Sweden. Specifically, when and why do teachers code switch in their questions? Is the language choice pedagogically motivated? Does the teachers' language choice have an influence over the students' language choice in their answers? The data were collected from three beginner-level Italian lessons in two Swedish High Schools. Transcripts of questions in which CS occurred were analyzed using a conversation analytic approach with a focus on sequentiality in relation to the organization of turn allocation, to understand when and why the CS occurs in both questions and answers. In our data teachers use CS when posing questions to the class in two major occasions, namely when the question is followed by a significant silence but also when no discernible silence follows a question. Furthermore, regardless of the language used by the teacher, the students answer in Italian when the questions are task-based, and in Swedish when the questions are off-task.
[ "Code-Switching", "Multilinguality" ]
[ 7, 0 ]
SCOPUS_ID:85020502303
"Clever tautologies"
This article covers a particular form of Wisława Szymborska conceptism, her Thought fugues, which turns the world's elements upside down and deconstructs them, uncovering their antinomies. The described reality takes the form of a concatenation, it is elliptic, based on seemingly symmetric tessellations, but in essence triggers some logic turns. It is possible to discern a similarity with advertising rhetoric and its constant variations of sense based on oppositions, such as double meaning, antiphrasis, paradox, or tautology. In all of which Szymborska is able to notice the condensing of meaning, or even some oxymoronic properties. It is perhaps this reason why her poetry is accessible to such a wide audience across the world. Though frequently condemned by 20th-century thinkers as a mere pleonasm, tautology becomes for the poetess a fluid material useful for her dialectics of imagination. Szymborska does not see language as a mirror of the world, but rather as an act of creating new worlds and new forms of life, like Wittgenstein did in his theory of language games. This is why she prefers such word games more than other poets. Due to their dual nature, tautologies constitute a potential and permanent element of Szymborska's conceptual games.
[ "Linguistics & Cognitive NLP", "Linguistic Theories" ]
[ 48, 57 ]
SCOPUS_ID:85048532446
"Climate Change" Frames Detection and Categorization Based on Generalized Concepts
The subliminal impact of framing of social, political and environmental issues such as climate change has been studied for a long time in political science and communications research. Media framing offers "interpretative package" for average citizens on how to make sense of climate change and its consequences to their livelihoods, how to deal with its negative impacts, and which mitigation or adaptation policies to support. A line of related work has used bag of words and word-level features to detect frames automatically in text. Such works face limitations since standard keyword based features may not generalize well to accommodate surface variations in text when different keywords are used for similar concepts. In this paper, we develop a new type of textual features that generalize (subject,verb,object) triplets extracted from text, by clustering them into high-level concepts. We utilize these concepts as features to detect frames in text. Our corpus comprises more than 45,000 climate change related sentences. Expert coders annotated those sentences as Frame/Non-Frame and framed sentences were mapped into one of four general frame categories: Solution, problem threat, cause, and motivation. Compared to uni-gram and bi-gram based models, classification and clustering using our generalized concepts yielded better discriminating features and a higher accuracy classifier with a 12% boost (i.e. from 74% to 83% in f-measure) and 0.91 clustering purity for Frame/Non-Frame detection.
[ "Information Extraction & Text Mining", "Information Retrieval", "Text Classification", "Text Clustering" ]
[ 3, 24, 36, 29 ]
SCOPUS_ID:85115936634
"Confine your women!": Diachronic development of Islamic interpretive discourse on the creation of woman
This article explores the diachronic development of Islamic interpretive discourse on the Qur'anic passage khalaqakum min nafsin waHidatin wa-khalaqa minha zawjaha, present in the first verse of Surat al-Nisa' and conventionally understood as the creation of the primeval couple, Adam and Eve. The analyses, performed within a theoretical framework of feminist discourse analysis, focus on ten medieval Sunni commentaries (tafasir) from the late third/ninth to the ninth/fifteenth centuries. The study reveals that the concept of nafs waHida, single soul, was interpreted as the first man, Adam, and the mate created from this soul, zawj, as Eve, the latter being created from the former's rib in all the exegetic accounts examined. These elaborated exegetic suppositions on human creation were strengthened throughout the classical period of tafsir. Interpretive information both accumulated and transformed in Islamic interpretive tradition through three discursive stages, characterised as normativisation, consolidation, and expanding the concept.
[ "Discourse & Pragmatics", "Explainability & Interpretability in NLP", "Semantic Text Processing", "Responsible & Trustworthy NLP" ]
[ 71, 81, 72, 4 ]
SCOPUS_ID:85019124272
"Contesting the cynicism of neoliberalism": A corpus-assisted discourse study of press representations of the Sino-US currency dispute
This article aims to expose the hegemony of neoliberalism in media discourse through a corpus-assisted discourse study of the representations of the Sino-US currency dispute in two newspapers - China Daily (CD) from China and The New York Times (NYT) from the US. The findings suggest that while neoliberal ideology can be identified in both CD and NYT, it is articulated and appropriated differently in the two newspapers to construct their respective stance towards the issue. Neoliberal beliefs are found pervading different levels of discourse (i.e., thematic, lexical and grammatical) in NYT to construct a combative stance towards China's exchange rate policies. However, the hegemony of neoliberalism can also be detected through CD's ambivalent stance towards change and the seemingly contradictory evaluation of the impact of exchange rate changes.
[ "Discourse & Pragmatics", "Semantic Text Processing", "Representation Learning" ]
[ 71, 72, 12 ]
SCOPUS_ID:84922722897
"Contextually" speaking: A survey of pragmatic learning abroad, in class, and online
In order to acquire pragmatic competence, learners must have access to the target language input and opportunities for pragmatic practice. Over the last three decades, research has emerged to specify this fundamental condition of pragmatic learning. Existing studies fall primarily into three main categories: study abroad literature that focuses on students' learning pragmatics in the target language community, formal classroom environment where pragmatics is not the target of instruction, and digitally-mediated contexts in which communication takes place in virtual environments. This paper synthesizes key findings in these three contexts, and compares and contrasts the opportunities and challenges involved in each context, with the overall aim of revealing how each context supports pragmatic learning and development.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85126070295
"Correlation between context and language at a high delicate level: taking Chinese emotion ? as an example"
In systemic functional linguistics, a delicate system of context is necessary for providing specific and checkable criteria for language choice. However, due to the complexity and vastness of the context system, little work has been undertaken for its extension, which results in a general and loose interpretation of the correlation between context and language. This study seeks to fill this gap by anchoring itself at a specific contextual domain concerning social conventions on emotion, namely the system of emotion. The correlation is addressed from two perspectives: a top-down perspective to explore how three crucial emotional features are realized via the selections from two linguistic systems, namely speech function and appraisal, and a bottom-up perspective to reveal the value construed by the realizational patterns for each feature. From a selected data corpus of modern Chinese novels, 8 realizational patterns are found, based on which the value systems of the emotional features are constructed; the Resonance Hypothesis is proposed to explain the process of correlation. The findings suggest that a delicate context system can explicitly demonstrate the correlation between context and language, thus theoretically strengthening their tie.
[ "Emotion Analysis", "Sentiment Analysis" ]
[ 61, 78 ]
http://arxiv.org/abs/2210.13039v1
"Covid vaccine is against Covid but Oxford vaccine is made at Oxford!" Semantic Interpretation of Proper Noun Compounds
Proper noun compounds, e.g., "Covid vaccine", convey information in a succinct manner (a "Covid vaccine" is a "vaccine that immunizes against the Covid disease"). These are commonly used in short-form domains, such as news headlines, but are largely ignored in information-seeking applications. To address this limitation, we release a new manually annotated dataset, ProNCI, consisting of 22.5K proper noun compounds along with their free-form semantic interpretations. ProNCI is 60 times larger than prior noun compound datasets and also includes non-compositional examples, which have not been previously explored. We experiment with various neural models for automatically generating the semantic interpretations from proper noun compounds, ranging from few-shot prompting to supervised learning, with varying degrees of knowledge about the constituent nouns. We find that adding targeted knowledge, particularly about the common noun, results in performance gains of upto 2.8%. Finally, we integrate our model generated interpretations with an existing Open IE system and observe an 7.5% increase in yield at a precision of 85%. The dataset and code are available at https://github.com/dair-iitd/pronci.
[ "Explainability & Interpretability in NLP", "Responsible & Trustworthy NLP" ]
[ 81, 4 ]
SCOPUS_ID:79954576335
"Delinquent behaviour as a kind of body politics" against adult regulations - Young people's discourses in Hong Kong
Usually, discourses of young people, particularly for the voices of 'youth-at-disadvantage', are silenced no matter in the social construction of their experiences or in the social investigation of their behaviour. Without their voices, a discourse of individual deficits is usually constructed to talk about their delinquent behaviour. However, in the discourses of young people, they considered their problem behaviour as a kind of 'body politics' to work against adult regulations. Moreover, obvious gender differences were revealed. While female adolescents constructed their problem behaviour as a tool to rebel against the harsh parental control on their leisure time activities, their male counterparts referred it as an escape to avoid heavy conflicts with parents on their school-to-work situations. In the discourses of young people, what they could do in the face of the adult society was only to rebel or escape with their bodies. Thus, in view of young people's discourse construction of their experiences, gender specific and family based interventions are recommended to work with 'youth-at-disadvantage' in outreaching social work service in Hong Kong. © 2011 Elsevier Ltd.
[ "Discourse & Pragmatics", "Semantic Text Processing", "Speech & Audio in NLP", "Multimodality" ]
[ 71, 72, 70, 74 ]
SCOPUS_ID:85139996855
"Después tampoco no podía trabajar". Negation in Spanish spoken by Mapudungun-Spanish bilinguals
Studies about the Spanish variety spoken by Mapudungun-Spanish bilinguals or español mapuchizado have reported a series of divergent features that makes it different from Chilean Spanish. However, until now no studies about the domain of negation in that variety of Spanish have been done. This work aims to study negation in this language variety from the perspective of language contact and linguistic typology. The study is done from twenty interviews directed to bilingual Mapudungun-Spanish bilinguals from two Mapuche communities from La Araucanía region, Chile. The analysis of the interview evidence the use of negative constructions in which both clausal negation no and N-words such as tampoco and nunca occur in the clause in preverbal position and the N-word occurring before clausal negative marker, a type of construction not found in standard Spanish. Despite the systematic occurrence of this constructions in the corpus, their frequency is low and is not conventionalized. Possible explanations for these constructions found in a language contact scenario are discussed.
[ "Typology", "Syntactic Text Processing", "Multilinguality" ]
[ 45, 15, 0 ]
SCOPUS_ID:79957581284
"Dialogue" in cross-cultural perspective: Japanese, Korean, and Russian discourses
Dialogue has become a key, cultural term in global English. Pleas for its use and enactment are prominent in many spheres of international activity. Following earlier works, this article explores terms (or characters) and practices which relate to dialogue in three cultural discourses: Japanese, Korean, and Russian. Revealed for each are the distinctive goals being targeted, implicit moral rules for conduct, as well as the proper tone, mode, and interactional structure at play. The distinctive features in each discourse of dialogue are discussed, as well as common features. Cross-cultural knowledge of this kind can clarify and address vexing problems such as the cultural balancing of information and truth with relational concerns. © 2011 National Communication Association.
[ "Discourse & Pragmatics", "Natural Language Interfaces", "Semantic Text Processing", "Dialogue Systems & Conversational Agents" ]
[ 71, 11, 72, 38 ]
SCOPUS_ID:49349097380
"Did they really say that?" - The women of Wenatchee: Vulnerability, confessions, and linguistic analysis
This article analyzes the vulnerabilities of certain women in custodial interrogations in the United States through legal, linguistic, and discourse analysis. At special risk are the impoverished or working class, the ill educated or illiterate, those with disabilities, or those who are nonnative speakers of English. Placing custodial interrogation into its U.S. legal setting, the study examines linguistic and discourse aspects of Miranda-related decisions and confessions. Drawing on data from the 1990s "Wenatchee Sex Ring" investigation and trials, the study examines the language of four women arrested and prosecuted in the Wenatchee cases. The analysis addresses discourse-level aspects of the data, including topic management and threats during the interrogation; the production of narratives as well as statements of location, result, and duration; and the sexual language used in the confessions. The article concludes that discourse analysis may provide an additional check on rogue interrogations of vulnerable people and once again calls for consistent recording of custodial interrogations in the United States. © 2008 Sage Publications.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
http://arxiv.org/abs/2010.00310v3
"Did you really mean what you said?" : Sarcasm Detection in Hindi-English Code-Mixed Data using Bilingual Word Embeddings
With the increased use of social media platforms by people across the world, many new interesting NLP problems have come into existence. One such being the detection of sarcasm in the social media texts. We present a corpus of tweets for training custom word embeddings and a Hinglish dataset labelled for sarcasm detection. We propose a deep learning based approach to address the issue of sarcasm detection in Hindi-English code mixed tweets using bilingual word embeddings derived from FastText and Word2Vec approaches. We experimented with various deep learning models, including CNNs, LSTMs, Bi-directional LSTMs (with and without attention). We were able to outperform all state-of-the-art performances with our deep learning models, with attention based Bi-directional LSTMs giving the best performance exhibiting an accuracy of 78.49%.
[ "Language Models", "Semantic Text Processing", "Representation Learning", "Sentiment Analysis", "Stylistic Analysis", "Multilinguality" ]
[ 52, 72, 12, 78, 67, 0 ]
http://arxiv.org/abs/2206.15010v1
"Diversity and Uncertainty in Moderation" are the Key to Data Selection for Multilingual Few-shot Transfer
Few-shot transfer often shows substantial gain over zero-shot transfer~\cite{lauscher2020zero}, which is a practically useful trade-off between fully supervised and unsupervised learning approaches for multilingual pretrained model-based systems. This paper explores various strategies for selecting data for annotation that can result in a better few-shot transfer. The proposed approaches rely on multiple measures such as data entropy using $n$-gram language model, predictive entropy, and gradient embedding. We propose a loss embedding method for sequence labeling tasks, which induces diversity and uncertainty sampling similar to gradient embedding. The proposed data selection strategies are evaluated and compared for POS tagging, NER, and NLI tasks for up to 20 languages. Our experiments show that the gradient and loss embedding-based strategies consistently outperform random data selection baselines, with gains varying with the initial performance of the zero-shot transfer. Furthermore, the proposed method shows similar trends in improvement even when the model is fine-tuned using a lower proportion of the original task-specific labeled training data for zero-shot transfer.
[ "Language Models", "Low-Resource NLP", "Semantic Text Processing", "Representation Learning", "Responsible & Trustworthy NLP", "Multilinguality" ]
[ 52, 80, 72, 12, 4, 0 ]
SCOPUS_ID:85058422755
"Do You Remember Going to the Beach?": References to Internal States in Polish and American Mother-Preschooler Shared Narratives
The present study examined references to cognitive states and emotions in narratives produced by mothers and preschoolers (aged 3 or 5 years) in Polish and American families. Participants were 32 mother-child dyads from Poland and 32 mother-child dyads from the United States. The two samples were matched with regard to child age, child gender, maternal age, and maternal education. The mother-child dyads were asked to tell three personal narratives. The co-constructed narratives were coded for mother and child references to cognitive states and emotions. Polish mothers were found to include significantly more references to cognitive states in their narratives than American mothers. Results also revealed significant correlations between mothers' and children's references to cognitive states across both samples. Related to child development, 5-year-olds produced significantly more tokens in the narratives than 3-year-olds. This study shows that mothers' use of cognitive state terms in shared narratives with their young children differs across two Western cultural contexts. The results of this study are discussed with regard to two themes in developmental psycholinguistics: relations between maternal and child language use, and cross-cultural variation.
[ "Psycholinguistics", "Linguistics & Cognitive NLP" ]
[ 77, 48 ]
SCOPUS_ID:85014783462
"Do animals have accents?": Talking with agents in multi-party conversation
In this paper we unpack the use of conversational agents, or so-called intelligent personal assistants (IPAs), in multiparty conversation amongst a group of friends while they are socialising in a café. IPAs such as Siri or Google Now can be found on a large proportion of personal smartphones and tablets, and are promoted as 'natural language' interfaces. The question we pursue here is how they are actually drawn upon in conversational practice? In our work we examine the use of these IPAs in a mundane and common-place setting and employ an ethnomethodological perspective to draw out the character of the IPA-use in conversation. Additionally, we highlight a number of nuanced practicalities of their use in multi-party settings. By providing a depiction of the nature and methodical practice of their use, we are able to contribute our findings to the design of IPAs.
[ "Natural Language Interfaces", "Dialogue Systems & Conversational Agents" ]
[ 11, 38 ]
SCOPUS_ID:84904735639
"Do you understand?": An analysis of native and non-native EFL teachers' questioning patterns at a Taiwanese cram school
A number of young Taiwanese students take extra English lessons at cram schools where classes are taught in English by native speakers of English; however, not much has been studied in such settings in previous literature. This study investigated questioning patterns of four cram school teachers, two native and two non-native speakers of English, with high- and a low-level classes of young learners. Eight types of questions used by the teachers and students were identified from 12 video- and audio-recorded lessons. Questioning patterns were analyzed and compared in terms of the teacher's language background and the students' proficiency levels. The study found that the class level, but not the teacher's language background, influenced how the teachers formed questions. The teachers with high-level students used more communicative question types, while the teachers with low-level students preferred instructional questions. In addition, it was found that using the target language as the sole medium in the classroom did not guarantee a communicative learning environment. When instructional questions dominated the classroom discourse, the students became passive in the interaction. The study suggests that EFL teachers should monitor the functions and effects of their questioning techniques so as to facilitate genuine interaction, even with low-level EFL learners.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:84900414135
"Double contingency" and semiotics of protest: An interdisciplinary perspective
The purpose of this paper is to briefly compare Barthes's and Luhmann's conceptions of the nature of communication, together with Stirner's theory of language. The relation between protest and processes of communication is analyzed by means of the concepts of "social structure" and "double contingency". The relation between latent and manifest functions is inquired upon in relation with the analysis of meaning and communication. Consequences for the theory of enunciation are discussed by considering also the effects of the adoption of a procedural model of inference.
[ "Linguistics & Cognitive NLP", "Linguistic Theories" ]
[ 48, 57 ]
SCOPUS_ID:84923882840
"Dr. Detective": Combining gamification techniques and crowdsourcing to create a gold standard in medical text
This paper proposes a design for a gamified crowdsourcing workflow to extract annotation from medical text. Developed in the context of a general crowdsourcing platform, Dr. Detective is a game with a purpose that engages medical experts into solving annotation tasks on medical case reports, tailored to capture disagreement between annotators. It incorporates incentives such as learning features, to motivate a continuous involvement of the expert crowd. The game was designed to identify expressions valuable for training NLP tools, and interpret their relation in the context of medical diagnosing. In this way, we can resolve the main problem in gathering ground truth from experts - that the low inter-annotator agreement is typically caused by different interpretations of the text. We report on the results of a pilot study assessing the usefulness of this game. The results show that the quality of the annotations by the expert crowd are comparable to those of an NLP parser. Furthermore, we observed that allowing game users to access each others' answers increases agreement between annotators.
[ "Information Extraction & Text Mining" ]
[ 3 ]
SCOPUS_ID:85063550582
"E Maruthuvachi"-Information extraction framework for data about obstetrics and gynecology in Tamil
Technology is transforming the world from traditional into Artificial Intelligence. Human beings are adopting themselves into the change using Technology. India is famous for the name of unique traditional culture. The traditional culture protected people to do useful things. Especially for women, they were protected by the traditional culture to gain knowledge about maternity and gynecology. The target framework to extract the useful information from the raw documents. The extraction process extracts the NLP elements from the raw documents. The framework is developed using modified model of neural network language model (NNLM). The proposed model is evaluated with F-Test. The evaluation produces the good result for accuracy.
[ "Language Models", "Semantic Text Processing", "Information Extraction & Text Mining" ]
[ 52, 72, 3 ]
SCOPUS_ID:84921557811
"Edison" and "Russell"
This chapter presents the text of the conversion between a roboticist and a theoretical neurobiologist about the issue of emotion. The roboticist suggested that it would be useful to have a list of definitions of key terms on the subject of emotion that takes into account logical alternative views. The theoretical neurobiologist argued that there is a linguistic risk in defining terms such as emotions, and that even if definitions were established there would still be possible problems in achieving consensus.
[ "Emotion Analysis", "Sentiment Analysis" ]
[ 61, 78 ]
SCOPUS_ID:85130531200
"Elinor's Talking to Me!":Integrating Conversational AI into Children's Narrative Science Programming
Video programs are important, accessible educational resources for young children, especially those from an under-resourced backgrounds. These programs' potential can be amplified if children are allowed to socially interact with media characters during their video watching. This paper presents the design and empirical investigation of interactive science-focused videos in which the main character, powered by a conversational agent, engaged in contingent conversation with children by asking children questions and providing responsive feedback. We found that children actively interacted with the media character in the conversational videos and their parents spontaneously provided support in the process. We also found that the children who watched the conversational video performed better in the immediate, episode-specific science assessment compared to their peers who watched the broadcast, non-interactive version of the same episode. Several design implications are discussed for using conversational technologies to better support child active learning and parent involvement in video watching.
[ "Visual Data in NLP", "Programming Languages in NLP", "Natural Language Interfaces", "Dialogue Systems & Conversational Agents", "Multimodality" ]
[ 20, 55, 11, 38, 74 ]
SCOPUS_ID:10244252883
"English Only" and our struggle to understand the concept of language
In this paper, I examine the impact of three language theories, nomenclaturism, segregationalism, and integrationalism, on our view of language. Specifically, I examine "English Only" laws and the way in which the legal community has relied on language theories, implicitly and explicitly, in responding to these laws. The "English Only" issue illustrates that the elements of "a language" are diverse and highly contextualized, and, therefore, I propose that the principle of integrationalism is the best way to study the complex, multidimensional nature of language. © 2004 Elsevier Ltd. All rights reserved.
[ "Linguistics & Cognitive NLP", "Linguistic Theories" ]
[ 48, 57 ]
SCOPUS_ID:77649193255
"Filming in Progress": New spaces for multimodal designing
Global trends call for new research to investigate multimodal designing mediated by new technologies and the implications for classroom spaces. This article addresses the relationship between new technologies, students' multimodal designing, and the social production of classroom spaces. Multimodal semiotics and sociological principles are applied to a series of claymation movie-making lessons in an upper primary school in Australia. The analysis focuses on the social meanings embedded in the multimodal spaces of the classroom-dialogic, bodily, embodied, architectonic, and screen spaces. The findings demonstrate how the uses of new technologies and the students' multimodal learning were tied to important transformations of space. © 2009 Elsevier Inc. All rights reserved.
[ "Multimodality" ]
[ 74 ]
SCOPUS_ID:84931363724
"Fortune at the bottom of the Classifier Pyramid": A Novel approach to Human Activity Recognition
"Fortune at the Bottom of the Pyramid", penned by the noted economist Prof. C. K. Prahalad1, talked about the wealth creating potentials of entities at the bottom of the economic hierarchy. Taking the same cue into classifier research in Machine Learning, Naive Bayes classifier is often the beaten boy among its contemporaries. In this work, Naive Bayes (NB), is revisited by looking at heuristic ways of improving any skewed data bias, systematic and weighted magnitude errors; to address the problem of human activity recognition and propose the Improved Naive Bayesian Algorithm. The novelty of this work is in adjusting the human activity recognition problem as a special case of text classification.
[ "Information Retrieval", "Text Classification", "Information Extraction & Text Mining" ]
[ 24, 36, 3 ]
SCOPUS_ID:84875346119
"Friend Moments": A Discursive Study of Friendship
The current study adopted an interactional perspective toward friendship and directly explored how pairs of self-identified friends practice friendship within a research conversation. Twelve pairs of young adult friends were interviewed together about their friendship, and the transcripts were analyzed using discourse analysis. During the interviews, participants performed their friendship in particular identifiable moments (i.e., friend moments) by addressing each other directly and drawing on locally shared resources (i.e., a shared personally relevant history, context, and resources), positioning themselves as insiders (i.e., friends), while the interviewer took up a position as an outsider (i.e., a nonfriend or stranger), often by remaining silent. Exploring friend moments revealed the plausibility of viewing friendship as actively interactional and nontrivially relational. Friend moments represent a particular kind of friendship talk where speakers accomplish the task of doing friendship in interaction, in personally relevant and situationally appropriate ways. © 2013 Copyright Taylor & Francis Group, LLC.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:85143621971
"From first words to the first form": 100 years afterwards
The material of the research includes A. N. Gvozdev’s diary entries entitled "From first words to the first form" and the records of the ‘speech start’ of three XXIst-century children. Taking the data into account, the author attempts to compare the speech acquisition of A. N. Gvozdev’s son Zhenya that is recorded in the above-mentioned diary and is 100 years distant from us and the author’s almost daily observation of her grandchildren’s speech development. This comparison aims to identify the constants and variables of such development. The analysis confirms the idea suggested by modern scientists that studying the formation of children’s ability to think and speak is a "clue" to many mysteries of the origin and evolution of communication. The study is based on empirical observation and takes into consideration the psychological, gender, social, and philosophical sides of the issue. The paper examines the phonetic, lexical, and grammatical aspects of children’s speech. It is claimed that each child forms laws of his/her own phonetics, and that the choice of sounds in a word to convey thoughts is a certain deliberate compromise between what he/she can pronounce and how he/she can do it to be understood. Mastering the lexical aspect of speech is closely related to the problem of children’s comprehension of the changing picture of the world. Mastering grammatical categories reflects variations in the formation of grammatical laws in children. It is concluded that each child creates his/her language model. At the same time, the communication environment requires children to adjust the created model in compliance with the norms adopted in a particular language.
[ "Language Models", "Semantic Text Processing", "Speech & Audio in NLP", "Syntactic Text Processing", "Phonetics", "Multimodality" ]
[ 52, 72, 70, 15, 64, 74 ]
SCOPUS_ID:84979800825
"Garbled demography" or "Demographization of the social"? - A Foueaultian Discourse Analysis of German Demographic Change at the Beginning of the 21st Century
German discourses of demographic change are characterized by alarmist scenarios. Especially since the turn of the millennium, a growing amount of publications addresses population aging and shrinking by depicting mostly dystopian future scenarios. Allegedly inevitable consequences with fundamental relevance for society are often proclaimed in the media and social-scientific discourses. Although most demographers alert to the fact that population projections should not be interpreted as prognoses, they are often employed as irrefutable knowledge as well as camouflage for normative positions. Complex demographic measures are frequently misinterpreted by journalists, who consequently produce "garbled demography" (Teitelbaum 2004). However, the "demographization of the social" (Barlösius 2007) turns out to be more complex than a misunderstanding or a distortion of "neutral" scientific facts. Michel Foucault's works provide a framework of suitable complexity in order to analyze the depth-structures of both discourses and their interrelations. This paper will first describe relevant conditions of existence of demographic knowledge orders, their rules of formation, and discursive regularities in order to shed light on the demographic ontology of the present. Subsequently, these depth structures will be related to preliminary results of a discourse analysis of 2900 press articles from leading German newspapers and journals covering the period from 2000 to 2012. In conclusion, first contours of a recently emerging post-alarmist discourse will be outlined. © GESIS.
[ "Discourse & Pragmatics", "Explainability & Interpretability in NLP", "Semantic Text Processing", "Responsible & Trustworthy NLP" ]
[ 71, 81, 72, 4 ]
SCOPUS_ID:84958554164
"Girlfriend getaway" as a contested term: Discourse analysis
All-female leisure travel is a fast growing tourism market segment that is frequently called "girlfriend getaways." This study explored the meanings associated with the "girlfriend getaway" term, using discourse analysis to understand the ways women build significance, activities, identities, relationships, politics, connections, and sign systems and knowledge with respect to it. Eleven focus groups and 15 individual interviews were conducted with 83 American and Canadian women. The analysis revealed that "girlfriend getaway" is a term with contested and polysemous meanings. While some women found it to be adequate, accurate, cute, and reflective of their all-female tourist experiences, others described it as stereotypical, narrow/claustrophobic, "pink," inadequate, and unreflective of their experiences. At times, the same symbolic meanings attracted some women but alienated others. Thus, tourism marketers need to identify and engage with different strands within their female clientele to ensure that their strategies appropriately respond to various preferences and lifestyles.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:77149161459
"Global" discourses of democracy and an English city
In many contemporary polities, democracy is portrayed as a universal good, a democratic ideal appears to be spreading globally, its practice burgeoning; it seems to be appearing for the first time in some places and deepening in established democracies. Yet, when one looks for the concrete touch of democracy in one's own activities, groups, communities and nation it becomes elusive. I discuss this apparent contradiction in relation to discourse and a new "Area Forum" in the English city of Preston. The categories of 'global' and 'local', 'identity' and 'branding' prove useful in discussing the contradiction as situated in the English context. I suggest that this problem of democracy may be understood in terms of the ideological concept of 'democratism': the assumption that the status quo in England is unproblematically democratic whilst discursively closing off the possibility of genuine democratic progress. © John Benjamins Publishing Company.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
http://arxiv.org/abs/1903.05210v1
"Hang in There": Lexical and Visual Analysis to Identify Posts Warranting Empathetic Responses
In the past few years, social media has risen as a platform where people express and share personal incidences about abuse, violence and mental health issues. There is a need to pinpoint such posts and learn the kind of response expected. For this purpose, we understand the sentiment that a personal story elicits on different posts present on different social media sites, on the topics of abuse or mental health. In this paper, we propose a method supported by hand-crafted features to judge if the post requires an empathetic response. The model is trained upon posts from various web-pages and corresponding comments, on both the captions and the images. We were able to obtain 80% accuracy in tagging posts requiring empathetic responses.
[ "Visual Data in NLP", "Responsible & Trustworthy NLP", "Ethical NLP", "Multimodality" ]
[ 20, 4, 17, 74 ]
SCOPUS_ID:85104733245
"Haptic Verbs" in current political speeches - An approach within the framework of embodied cognition and enaction
This contribution shows that mental concepts are not static but fluid in nature and are constituted by a combination of perceptual parameters anchored in our visual perception, corporal experience and neuronal simulation. We will focus on the haptic verbs donner ('to give', volitional [transfer]), prendre ('to take', [transfer] of possession), porter ('to carry', [transfer] to a different location) and battre ('to hit', [transfer] of force) as well as on the fact that these verbs have great persuasive power not only in politics but also in daily life. Furthermore, we will try to show that haptic verbs are an essential linguistic tool for political power, which, despite its importance, has not yet been addressed in political discourse analysis. To do so, we will look at how the three candidates for the 2017 French presidential election, François Fillon, Marine Le Pen and Emmanuel Macron, made use of the aforementioned haptic verbs.
[ "Discourse & Pragmatics", "Semantic Text Processing", "Speech & Audio in NLP", "Multimodality" ]
[ 71, 72, 70, 74 ]
http://arxiv.org/abs/1709.01950v1
"Having 2 hours to write a paper is fun!": Detecting Sarcasm in Numerical Portions of Text
Sarcasm occurring due to the presence of numerical portions in text has been quoted as an error made by automatic sarcasm detection approaches in the past. We present a first study in detecting sarcasm in numbers, as in the case of the sentence 'Love waking up at 4 am'. We analyze the challenges of the problem, and present Rule-based, Machine Learning and Deep Learning approaches to detect sarcasm in numerical portions of text. Our Deep Learning approach outperforms four past works for sarcasm detection and Rule-based and Machine learning approaches on a dataset of tweets, obtaining an F1-score of 0.93. This shows that special attention to text containing numbers may be useful to improve state-of-the-art in sarcasm detection.
[ "Stylistic Analysis", "Sentiment Analysis" ]
[ 67, 78 ]
SCOPUS_ID:84936998631
"He just has to like ham" - The centrality of meat in home and consumer studies
This study aimed to describe Discourses on meat in the school subject Home and Consumer Studies in five different northern Swedish schools. Fifty-nine students and five teachers from five different schools were recorded and in some cases video-taped during lessons. Results indicate that meat was seen as central to nutritional health, sensory experience, culture and social relationships. This positive view was challenged by an alternative Discourse where meat was threatening to health, sensory experience and psychological comfort, but this was not strong enough to affect centrality. Even when participants sought to promote the health advantages of reducing meat consumption, the dominant centrality Discourse was strengthened. This implies that the possible tension between physical and psychosocial/emotional health can make the benefits of a reduction difficult both to convey and accept. A form of critical food literacy may help teachers deconstruct the arbitrary power of the centrality Discourse, but it may also strengthen meat-eater identities because the social norms that guide food choice become salient. A redesign of Discourses might facilitate a reduction in meat consumption, but such a paradigm shift is dependent on the development of society as a whole, and can only be briefly touched upon within the limited time frames and resources of Home and Consumer Studies.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:84884474843
"Heaven and Hell on Earth" A critical discourse analysis of religious terms in Norwegian autobiographies describing personal experience of mental health problems
This article explores the use of religious terms in six Norwegian autobiographies written between 1925 and 2005 by people who themselves have been patients in the mental health services. Through a critical discourse analysis, we discuss the functions of religious discourse in the texts and its position in contrast to the medical discourse predominant in today's mental health services. It was found that religious (predominantly Christian) terms were used to varying degrees in all autobiographies as a means to capture the immensity and inherent ambivalence characteristic of mental health problems. Despite the "medical turn" in professional mental health discourse, there is no clear evidence of a decrease in the use of religious terms from the oldest to the most recent text. We propose that professional mental health workers to a larger extent take into account the religious dimension in therapy, and reflect on its larger historical and sociocultural context. © 2013 Taylor and Francis Group, LLC.
[ "Discourse & Pragmatics", "Semantic Text Processing", "Ethical NLP", "Responsible & Trustworthy NLP" ]
[ 71, 72, 17, 4 ]
SCOPUS_ID:84878394208
"Help me, i need more user tests!" User simulations as supportive tool in the development process of spoken dialogue systems
In this paper we present our experiences in developing a spoken dialogue system supported by tests with a user simulation. Since the code of dialogue systems with modest complexity can easily get unclear, it is almost impossible to deliver error-free systems without user tests in the development process. We show how we included our user simulation environment SpeechEval in the development process of three VoiceXML dialogue systems and discuss advantages and drawbacks compared to tests with real users.
[ "Natural Language Interfaces", "Dialogue Systems & Conversational Agents" ]
[ 11, 38 ]
SCOPUS_ID:67349091172
"Here the scientists explain what i said." Coordination practices elicited during the enactment of the results and discussion sections of adapted primary literature0
Adapted primary literature (APL) is a novel text genre that retains the authentic characteristics of primary literature. Learning through APL represents an educational intervention with an authentic scientific context. In this case study, we analyzed the 80-min discourse developed during the enactment of an article from an APL-based curriculum in biotechnology in one class, and examined epistemic practices used by students during their meaning-making of the Results and Discussion sections of the article. Specifically, we examined coordination practices, by which students connected elements belonging to different epistemic status or context (theory, data, experimental stages, biotechnological applications and text). The application of coordination practices was identified more than 70 times during the lesson. In the context of the Results section, the students displayed research-oriented coordination practices, which were frequently associated with claims of comprehension difficulty. In the context of the Discussion section, students displayed text-oriented coordination practices, associated with analysis of the text characteristics. We are suggesting that the research-oriented coordination practices and some of the text-oriented ones enabled the emergence of authentic scientific practices and learning by inquiry. Another type of text-oriented coordination practice enabled reflection on scientists' experimental processes, enabling learning science as inquiry. The enactment model of APL used here allowed for both the emergence of the two dimensions of inquiry learning and the promotion of scientific literacy in the fundamental and derived senses. © Sringer Science+Business Media B.V. 2008.
[ "Discourse & Pragmatics", "Explainability & Interpretability in NLP", "Semantic Text Processing", "Responsible & Trustworthy NLP" ]
[ 71, 81, 72, 4 ]
SCOPUS_ID:84871046014
"Heteroglossia" in Czech folk songs: Some text-linguistic probes into their texture
This paper demonstrates the mutual relations between the various "voices" in Czech folk songs and the way in which these voices can be identified. The introductory section characterizes the specific features of the songs: the concision and condensation of their structure, as well as their emotionality of expression. The second section is devoted to the analysis of a selected corpus of folk songs from the perspective of heteroglossia. Their texture appears as a net of relations and functions which operate on different levels, stand in mutual complementarity, partly overlapping, and show differing relevance. Using methods of text linguistics, the author concentrates on two relations and functions - speech acts and compositional functions - as well as personal attitudes. The analysis focuses on utterances with the communicative function of appeal, i.e. on the interrogative and addressing turns which are quite frequent in Czech folk songs, on their interactive structure and modes of direct and indirect speech. The author shows that the song genre has two faces: it is a compact combination of verbal text and its musical part (the tune) which may appear to be uneven in their mutual relations. The last section deals with the phenomenon of intertextuality in the song structure.
[ "Speech & Audio in NLP", "Multimodality" ]
[ 70, 74 ]
SCOPUS_ID:85068771814
"Hey Google, do unicorns exist?": Conversational agents as a path to answers to children's questions
Children are known to be curious and persistent questionaskers. The pervasiveness of voice interfaces represents an opportunity for children who are not yet fluent readers to independently search the Internet by asking questions through conversational agents such as Amazon Alexa, Apple's Siri, and the Google Assistant. Through a twoweek, in-home deployment study involving 18 families (children aged 5-6 and their parents), we report on which questions children choose to ask the conversational agent, the answers the agent provided, challenges in use, and their perceptions of the technology. Based on our analysis, we identify several considerations for the design of voicebased conversational agents that aim to support young children's question-asking behavior and subsequent development.
[ "Natural Language Interfaces", "Multimodality", "Speech & Audio in NLP", "Dialogue Systems & Conversational Agents" ]
[ 11, 74, 70, 38 ]
SCOPUS_ID:3042511850
"Hire" or "Fire"? Taking AD-vantage of innovations in the Japanese syllabary system
[ "Phonology", "Syntactic Text Processing" ]
[ 6, 15 ]
SCOPUS_ID:85117251604
"Hope it's not just the honeymoon phase": online discursive portrayals of migrant domestic helpers
This paper examines online discursive representations of migrant domestic helpers (MDHs) by Hong Kong employers. Unlike existing research, which concentrates on the experiences of MDHs from their own perspectives, this study focuses on positive narrations about MDHs by their employers. Using critical discourse analysis, this study identified the discursive strategies deployed to portray MDHs in more than 2,000 Facebook posts. The findings reveal that, although the interlocutors attempted to commend the MDHs in their employ, they also emphasised their own superiority by portraying themselves as gastronomic experts, good educators, and benefactors, thus developing an ideological paradox. Another dimension of ideological ambivalence concerned the discursive conflict between their high expectations from the MDHs and their underlying belief that domestic work neither requires skills nor deserves high pay. Taken together, these factors are responsible for the entrenchment of the inferior image of MDHs in Hong Kong society, despite the persistent endeavours of activist groups to spread awareness of their exploitation.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
http://arxiv.org/abs/2109.13489v1
"How Robust r u?": Evaluating Task-Oriented Dialogue Systems on Spoken Conversations
Most prior work in dialogue modeling has been on written conversations mostly because of existing data sets. However, written dialogues are not sufficient to fully capture the nature of spoken conversations as well as the potential speech recognition errors in practical spoken dialogue systems. This work presents a new benchmark on spoken task-oriented conversations, which is intended to study multi-domain dialogue state tracking and knowledge-grounded dialogue modeling. We report that the existing state-of-the-art models trained on written conversations are not performing well on our spoken data, as expected. Furthermore, we observe improvements in task performances when leveraging n-best speech recognition hypotheses such as by combining predictions based on individual hypotheses. Our data set enables speech-based benchmarking of task-oriented dialogue systems.
[ "Speech & Audio in NLP", "Robustness in NLP", "Natural Language Interfaces", "Text Generation", "Dialogue Systems & Conversational Agents", "Responsible & Trustworthy NLP", "Speech Recognition", "Multimodality" ]
[ 70, 58, 11, 47, 38, 4, 10, 74 ]
SCOPUS_ID:85146199255
"How Was the Match?": Semantic Similarity between Electronic Media Commentary and Work Domain Analysis Key Phrases
Football player's performance can be measured in an objective way (e. g. Goals scored, assists, interceptions), this being seldom a method to compare and rank the best players by categories. Over years of study, many other factors that can influence the players performance were discovered and studied, considering not only objective factors, but also subjective factors. Match commentary from different sources (e.g., social and formal media) also plays an important role on a more subjective performance assessment. By using semantic similarity analysis, this study aims to contribute to the understanding of the concepts that are used in this commentaries, notably to each extend key phrases associated to match processes are used in commentaries published in social and formal media.
[ "Semantic Text Processing", "Semantic Similarity" ]
[ 72, 53 ]
SCOPUS_ID:10444274177
"How can you tell?" Towards a common sense explanation of conversational code-switching
Rational Choice (RC) models of code-switching argue that bilingual speakers make rational choices according to the rights and obligations they perceive in a given situation. Some situations are marked and some unmarked. Speakers choose their languages to index their rational decisions, as well as their attitudes and identities. The Conversation Analysis (CA) approach to code-switching agrees with the RC model that bilingual speakers are rational individuals. But instead of being oriented to rights and obligations, or attitudes and identities, bilingual speakers are first and foremost assumed to be oriented to conversational structures aiming primarily at achieving coherence in the interactional task at hand. Their language choice and code-switching is therefore 'programmatically relevant' to the talk-in-interaction. The CA approach therefore begins where the RC model stops and seeks evidence from talk-in-interaction rather than from external knowledge of community structure and relations. Examples of conversational code-switching by Chinese-English bilinguals will be cited to support the arguments. © 2004 Elsevier B.V. All rights reserved.
[ "Code-Switching", "Commonsense Reasoning", "Explainability & Interpretability in NLP", "Natural Language Interfaces", "Dialogue Systems & Conversational Agents", "Responsible & Trustworthy NLP", "Reasoning", "Multilinguality" ]
[ 7, 62, 81, 11, 38, 4, 8, 0 ]
SCOPUS_ID:43249141105
"How come most people don't see it?": Slashing the Lord of the rings
The now well-established fan tradition of "slash fiction" locates homoerotic undercurrents beneath the surface of popular films, television serials, and books, from Star Trek to Pride and Prejudice. The encoding/decoding model of media production and reception has recently been used to explain how enthusiasts of slash fiction are able to discern subtexts invisible to the majority of readers and viewers, with those enthusiasts' discussions of texts being cited as evidence; here, it is argued that this mis-characterises complex rhetorical manoeuvres as transparent reports on private comprehension processes. A sample of online fan discourse regarding one particular homoerotic pairing is analysed, it being proposed that reception study as a whole must re-conceptualise the data upon which it most heavily relies; namely, spoken or written reports of encounters with texts. This forms part of an ongoing project employing discursive psychology and the study of argumentation to investigate reading and textual culture. © 2007 Taylor & Francis.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:35648995839
"How may I help you?" Politeness in computer-mediated and face-to-face library reference transactions
Drawing upon Wittgenstein's theory of language games, we apply a pragmatic approach to organizational communication. The study extends current theories of computer-mediated communication, and explores how language actually gets used across media. We conduct a comparative study of face-to-face versus computer-mediated reference transactions in an academic library, and analyze people's use of politeness strategies. The study observes that people use politeness strategies to play different language games across media, and that a dynamic interplay exists among the three constituents of language games: sense making, language use, and forms of life. In particular, we found that: First, people use significantly more negative politeness strategies and fewer positive politeness strategies online than face-to-face. Second, language use influences people's understanding and precedes practical forms of life. Third, CMC is a different form of life than a face-to-face communication. The new online form of life shapes people's sense making and the way they use language, resulting in emerging new grammars of CMC. Finally, our findings suggest the interplay among multiple forms of life. This study offers important theoretical and managerial implications for organizational communication and the production and delivery of services in the rapidly expanding digital economy. © 2007 Elsevier Ltd. All rights reserved.
[ "Linguistics & Cognitive NLP", "Linguistic Theories" ]
[ 48, 57 ]
SCOPUS_ID:84922041868
"How much context do you need?" An experiment about the context size in Interactive Cross-language Question Answering.
The main topic of this paper is the context size needed for an efficient Interactive Cross-language Question Answering system. We compare two approaches: the first one (baseline system) shows user whole passages (maximum context: 10 sentences). The second one (experimental system) shows only a clause (minimum context). As cross-language system, the main problem is that the language of the question (Spanish) and the language of the answer context (English) are different. The results show that large context is better. However, there are specific relations between the context size and the knowledge about the language of the answer: users with poor level of English prefer context with few words.
[ "Natural Language Interfaces", "Question Answering" ]
[ 11, 27 ]
SCOPUS_ID:33749610800
"How much context do you need?": An experiment about the context size in Interactive Cross-Language Question Answering
The main topic of this paper is the context size needed for an efficient Interactive Cross-language Question Answering system. We compare two approaches: the first one (baseline system) shows the user whole passages (maximum context: 10 sentences). The second one (experimental system) shows only a clause (minimum context). As cross-language system, the main problem is that the language of the question (Spanish) and the language of the answer context (English) are different. The results show that large context is better. However, there are specific relations between the context size and the knowledge about the language of the answer: users with poor level of English prefer context with few words. © Springer-Verlag Berlin Heidelberg 2006.
[ "Passage Retrieval", "Natural Language Interfaces", "Question Answering", "Information Retrieval" ]
[ 66, 11, 27, 24 ]
SCOPUS_ID:84887731559
"How-to" questions answering using relations-based summarization
The problem considered in this paper relates to searching for "How-to" question answers and identifying the main semantic part in the answers found. We propose a "bag-of-relations" method for document summarization. This approach consists in identifying sentences that correspond to the key relations the most. Rating of each relation in the training set of question answering documents is evaluated on the basis of its frequency, as well as on the basis of semantic closeness to other relations included in the set. We propose a method allowing automatic selection of documents for the training set and for refining the results of a full-text search. We also suggest a method of evaluating the quality of the document containing an answer to a "Howto" question, as well as a method of identifying main objects required to perform the actions described in the document found. © IDOSI Publications, 2013.
[ "Question Answering", "Summarization", "Natural Language Interfaces", "Text Generation", "Information Extraction & Text Mining" ]
[ 27, 30, 11, 47, 3 ]
SCOPUS_ID:85091269022
"Human, All Too Human": NOAA Weather Radio and the Emotional Impact of Synthetic Voices
The integration of text-to-speech into an open technology stack for low-power FM community radio stations is an opportunity to automate laborious processes and increase accessibility to information in remote communities. However, there are open questions as to the perceived contrast of synthetic voices with the local and intimate format of community radio. This paper presents an exploratory focus group on the topic, followed by a thematic analysis of public comments on YouTube videos of the synthetic voices used for broadcasting by National Oceanic and Atmospheric Administration (NOAA) Weather Radio. We find that despite observed reservations about the suitability of TTS for radio, there is significant evidence of anthropomorphism, nostalgia and emotional connection in relation to these voices. Additionally, introduction of a more "human sounding" synthetic voice elicited significant negative feedback. We identify pronunciation, speed, suitability to content and acknowledgment of limitations as more relevant factors in listeners' stated sense of connection.
[ "Speech & Audio in NLP", "Multimodality" ]
[ 70, 74 ]
SCOPUS_ID:85129811319
"I Have No Text in My Post": Using Visual Hints to Model User Emotions in Social Media
As an emotion plays an important role in people's everyday lives and is often mirrored in their social media use, extensive research has been conducted to characterize and model emotions from social media data. However, prior research has not sufficiently considered trends of social media use - the increasing use of images and the decreasing use of text - nor identified the features of images in social media that are likely to be different from those in non-social media. Our study aims to fill this gap by (1) considering the notion of visual hints that depict contextual information of images, (2) presenting their characteristics in positive or negative emotions, and (3) demonstrating their effectiveness in emotion prediction modeling through an in-depth analysis of their relationship with the text in the same posts. The results of our experiments showed that our visual hint-based model achieved 20% improvement in emotion prediction, compared with the baseline. In particular, the performance of our model was comparable with that of the text-based model, highlighting not only a strong relationship between visual hints of the image and emotion, but also the potential of using only images for emotion prediction which well reflects current and future trends of social media use.
[ "Visual Data in NLP", "Emotion Analysis", "Sentiment Analysis", "Multimodality" ]
[ 20, 61, 78, 74 ]
SCOPUS_ID:85085639330
"I Hear You, i Feel You": Encouraging Deep Self-disclosure through a Chatbot
Chatbots have great potential to serve as a low-cost, effective tool to support people's self-disclosure. Prior work has shown that reciprocity occurs in human-machine dialog; however, whether reciprocity can be leveraged to promote and sustain deep self-disclosure over time has not been systematically studied. In this work, we design, implement and evaluate a chatbot that has self-disclosure features when it performs small talk with people. We ran a study with 47 participants and divided them into three groups to use different chatting styles of the chatbot for three weeks. We found that chatbot self-disclosure had a reciprocal effect on promoting deeper participant self-disclosure that lasted over the study period, in which the other chat styles without self-disclosure features failed to deliver. Chatbot self-disclosure also had a positive effect on improving participants' perceived intimacy and enjoyment over the study period. Finally, we reflect on the design implications of chatbots where deep self-disclosure is needed over time.
[ "Natural Language Interfaces", "Dialogue Systems & Conversational Agents" ]
[ 11, 38 ]
SCOPUS_ID:84873940602
"I Wouldn't Have Said it That Way": Mediating professional editorial comments in a secondary school science classroom
This article presents an analysis of a videotaped lecture from a secondary school science classroom. The students in this class had drafted science journalism articles and submitted them for professional editorial review and possible publication in a science newsmagazine for a teenage audience. Before allowing her students to see the editorial feedback, the teacher prepared a lecture along with a handout; she appeared to be trying to mediate the editor's comments. Using discourse analysis, this article examines the teacher's varied roles in this class period as she worked to create a hybrid space, a space that was not a typical science classroom but was also not a professional newsroom. The article concludes that this teacher's repositioning was necessary in order for her to open her classroom to more authentic learning opportunities for her students. © 2013 Elsevier Inc.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:84919624765
"I am a waste of breath, of space, of time": Metaphors of self in a pro-anorexia group
According to recent research on eating disorders, heavy users of pro-anorexia (pro-ana) sites show higher levels of disordered eating and more severe impairment of quality of life than non-heavy users. A better understanding of how pro-ana members self-present in the virtual world could shed some light on these offline behaviors. Through discourse analysis, I examined the metaphors the members of a pro-ana group invoked in their personal profiles on a popular social networking site, to talk about the self. I applied the Metaphor Identification Procedure to 757 text profiles. I identified four key metaphorical constructions in pro-ana members' self-descriptions: self as space, self as weight, perfecting the self, and the social self. These four main metaphors represented discourse strategies, both to create a collective pro-ana identity and to enact an individual identity as pro-ana. In this article, I discuss the implications of these findings for the treatment of eating disorders.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]
SCOPUS_ID:22844438431
"I am not a drug abuser, I am a drug user": A discourse analysis of 44 drug users' construction of identity
Based on individual conversational interviews with 44 socially integrated drug users in Stockholm, this article examines the informants' self-presentations and their representations of drug abusers. The results show that the informants strive towards positive self-presentation. In this process, the drug abuser identity is important as it provides a negative identity that reinforces the informants' desired self-presentation. The mechanisms of negative Other-presentations derive from a background in which the informants attempt to escape a socially ascribed deviant identity and exchange it for a not yet stabilised positive identity. © 2005 Taylor & Francis Group Ltd.
[ "Discourse & Pragmatics", "Semantic Text Processing" ]
[ 71, 72 ]