Guilhem Santé commited on
Commit
6e26f2e
·
unverified ·
1 Parent(s): 1527eb0

docs: description for dataset card

Browse files
Files changed (1) hide show
  1. README.md +35 -1
README.md CHANGED
@@ -35,4 +35,38 @@ tags:
35
  pretty_name: liaisons IBM's Claim Stance Dataset Sample
36
  size_categories:
37
  - n<1K
38
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
35
  pretty_name: liaisons IBM's Claim Stance Dataset Sample
36
  size_categories:
37
  - n<1K
38
+ ---
39
+ ---
40
+ # Dataset Card for Liaison's IBM Claim Stance Dataset Sample
41
+
42
+ ## Table of Contents
43
+
44
+ - [Dataset Summary](#dataset-summary)
45
+ - [Dataset Structure](#dataset-structure)
46
+ - [Licensing Information](#licensing-information)
47
+ - [Citation Information](#citation-information)
48
+
49
+ ## Dataset Summary
50
+
51
+ The present dataset is a result of processing the [IBM Debater Claim Stance Dataset](https://huggingface.co/datasets/ibm/claim_stance) to create representative samples. The size has been reduced to roughly 100 entries, enabling the benchmarking of models for relation-based argument mining tasks with limited resources.
52
+
53
+ The sample also modifies the original dataset to achieve a more balanced plurality of stances and topics, and creates a new "unrelated" class in argument relation (following a simple rule-based data augmentation algorithm). Further details on the preprocessing can be found on [GitHub](https://github.com/coding-kelps/liaisons-preprocess).
54
+
55
+ ## Dataset Structure
56
+
57
+ * parent_argument - The first argument that states a position regarding a topic
58
+ * child_argument - Another argument that is compared to the parent argument
59
+ * relation - The argumentative relation of the child argument to the parent argument. It can either be support/attack in the binary split or support/attack/unrelated in the ternary split
60
+
61
+ ## Licensing Information
62
+
63
+ This work includes data from the following sources:
64
+
65
+ * Wikipedia content licensed under CC BY-SA 3.0: [Wikipedia](https://en.wikipedia.org/wiki/Wikipedia:Copyrights#Reusers.27_rights_and_obligations)
66
+ * IBM content licensed under CC BY-SA 3.0: (c) Copyright IBM 2014. Released under [CC-BY-SA 3.0](http://creativecommons.org/licenses/by-sa/3.0/)
67
+
68
+ Modifications and preprocessing have been made to the original data. This derivative work is licensed under the same CC BY-SA 3.0 license.
69
+
70
+ ## Citation Information
71
+
72
+ Further information about the original dataset can be found on its original [HuggingFace page](https://huggingface.co/datasets/ibm/claim_stance) and its associated research papers: [Stance Classification of Context-Dependent Claims](https://aclanthology.org/E17-1024/) and [Improving Claim Stance Classification with Lexical Knowledge Expansion and Context Utilization](https://aclanthology.org/W17-5104/).