Update README.md
Browse files
README.md
CHANGED
@@ -46,7 +46,7 @@ pretty_name: Multi^3Hate
|
|
46 |
size_categories:
|
47 |
- 1K<n<10K
|
48 |
---
|
49 |
-
# Multi<sup>3</sup>Hate
|
50 |
|
51 |
<!-- Provide a quick summary of the dataset. -->
|
52 |
|
@@ -66,8 +66,22 @@ finding that these models align more closely with annotations from the US than w
|
|
66 |
|
67 |
## Dataset Details
|
68 |
|
|
|
|
|
|
|
|
|
|
|
69 |
|
70 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
71 |
|
72 |
<!-- Provide the basic links for the dataset. -->
|
73 |
|
@@ -76,24 +90,28 @@ finding that these models align more closely with annotations from the US than w
|
|
76 |
|
77 |
|
78 |
|
79 |
-
####
|
80 |
|
81 |
-
<!-- This section describes the
|
82 |
|
83 |
-
|
84 |
|
85 |
-
|
86 |
|
87 |
-
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
|
88 |
|
89 |
-
|
|
|
|
|
|
|
|
|
|
|
90 |
|
91 |
## License
|
92 |
|
93 |
CC BY-NC-ND 4.0
|
94 |
|
95 |
|
96 |
-
## Citation
|
97 |
|
98 |
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
|
99 |
|
|
|
46 |
size_categories:
|
47 |
- 1K<n<10K
|
48 |
---
|
49 |
+
# Multi<sup>3</sup>Hate Dataset: Multimodal, Multilingual, and Multicultural Hate Speech Dataset
|
50 |
|
51 |
<!-- Provide a quick summary of the dataset. -->
|
52 |
|
|
|
66 |
|
67 |
## Dataset Details
|
68 |
|
69 |
+
The dataset comprises a curated collection of 300 memes—images paired with embedded captions—a prevalent form of multimodal content,
|
70 |
+
presented in five languages: English (en), German (de), Spanish (es), Hindi (hi), and Mandarin (zh).
|
71 |
+
Each of the 1,500 memes (300×5 languages) is annotated for hate speech in the respective target language by at least five native speakers
|
72 |
+
from the same country. These countries were chosen based on the largest number of native speakers of each target language: USA (US), Germany (DE),
|
73 |
+
Mexico (MX), India (IN), and China (CN).
|
74 |
|
75 |
+
|
76 |
+
We curate a list of culturally relevant templates by filtering them according to sociopolitical categories.
|
77 |
+
These categories were discussed and decided among the authors. Each category is further divided into specific topics
|
78 |
+
based on established criteria.
|
79 |
+
|
80 |
+
<img src="https://cdn-uploads.huggingface.co/production/uploads/65e0e6fa4394fc3d1b59627a/mBi8QLRazQ2glLaC12qMm.png" width="500"/>
|
81 |
+
|
82 |
+
|
83 |
+
|
84 |
+
### Dataset Sources
|
85 |
|
86 |
<!-- Provide the basic links for the dataset. -->
|
87 |
|
|
|
90 |
|
91 |
|
92 |
|
93 |
+
#### Annotation process & Demographic
|
94 |
|
95 |
+
<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
|
96 |
|
97 |
+
We refer to our paper for more details.
|
98 |
|
99 |
+
<img src="https://cdn-uploads.huggingface.co/production/uploads/65e0e6fa4394fc3d1b59627a/3egu-bS9rF4CM5zJcCJoi.png" width="500"/>
|
100 |
|
|
|
101 |
|
102 |
+
|
103 |
+
#### Who are the source data producers?
|
104 |
+
|
105 |
+
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
|
106 |
+
|
107 |
+
The original meme images were crawled from https://memegenerator.net
|
108 |
|
109 |
## License
|
110 |
|
111 |
CC BY-NC-ND 4.0
|
112 |
|
113 |
|
114 |
+
## Citation
|
115 |
|
116 |
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
|
117 |
|