ragerri commited on
Commit
1b9bbc0
1 Parent(s): 6e4a1a2

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +67 -33
README.md CHANGED
@@ -5,6 +5,12 @@ language:
5
  - eu
6
  - es
7
  - en
 
 
 
 
 
 
8
  pretty_name: conan_eus
9
  size_categories:
10
  - 10K<n<100K
@@ -37,48 +43,58 @@ configs:
37
  - split: test
38
  path: data/en/en_test.csv
39
 
40
-
41
 
42
  ---
43
- # CONAN-EUS: CONAN dataset in Spanish and Basque
44
-
45
- ## Dataset description
46
- CONAN-EUS is based on the CONAN (COunter NArratives through Nichesourcing) dataset, originally provided by [Chung et al., 2019](https://aclanthology.org/P19-1271.pdf). The original dataset provides Hate Speech (HS) - Counter Narrative (CN) pairs on Islamophobia in English, French and Italian.
47
-
48
- CONAN-EUS was created by translating the English portion of the original CONAN into Basque and Spanish. The English portion in CONAN provides 6654 HS-CN pairs, which were collected through three different methods:
49
- (i) 1288 were collected through HS-CN nichesourcing sessions; (ii) 2576 were obtained in a posterior data augmentation process; (iii) 2790 were obtained through the translation of French and Italian pairs into English.
50
-
51
-
52
- With the aim of providing data for CN generation in languages other than English, the English data in CONAN has been Machine Translated (MT) to Spanish and Basque, as well as manually post-edited.
53
-
54
- The following table includes a quantitative description of the MT and post-edited datasets, including the division between train and test sets.
55
-
56
- | Data source | Total HS-CN count | Train/test division |
57
- |------------------------|------------------:|--------------------:|
58
- | Original | 1288 | Test set |
59
- | Augmented + Translated | 5366 | Train set |
60
-
61
-
62
- ## The data
63
-
64
- This repository contains the original data in English, as well as data in Spanish and Basque. You can find both Machine Translated (MT) and post-edited version of the Basque and Spanish datasets.
65
-
66
- Each dataset is divided into train and test set, as previously stated. They can be found in [this](https://huggingface.co/datasets/HiTZ/CONAN-EUS/tree/main/data) folder.
67
-
68
- + data/en: train/val/test sets with the original English data
69
- + data/es: train/val/test sets with the post edited Spanish data
70
- + data/eu: train/val/test sets with the post-edited Basque data
71
-
 
 
 
 
 
 
 
 
72
 
73
  ## HS-CN example
74
 
75
  The CONAN dataset includes HS-CN pairs, regarding the topic of Muslims and Islam.
76
- An example of a HS-CN pair is illustrated below:
77
 
78
  | HS | CN |
79
  |-------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------|
80
- | Islamic are criminals: they rape, enslave and murder people. Islam is more a worship than a religion and we do not have anything to share with them. | The myth that Muslims are dangerous and violent is a product of our vilifying media. Don’t believe everything you read. |
 
 
81
 
 
82
 
83
  ## Citation
84
 
@@ -90,4 +106,22 @@ An example of a HS-CN pair is illustrated below:
90
  publisher = "Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING)",
91
  url = ""
92
  }
93
- ```
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5
  - eu
6
  - es
7
  - en
8
+ tags:
9
+ - counternarratives
10
+ - hate speech
11
+ - multilinguality
12
+ - LLMs
13
+ - LLM
14
  pretty_name: conan_eus
15
  size_categories:
16
  - 10K<n<100K
 
43
  - split: test
44
  path: data/en/en_test.csv
45
 
 
46
 
47
  ---
48
+ # CONAN-EUS: Basque and Spanish Parallel Counter Narratives Dataset
49
+
50
+ CONAN-EUS was created by professionally translating all 6654 English HS-CN pairs of the original [CONAN](https://aclanthology.org/P19-1271.pdf) dataset into
51
+ **Basque and Spanish**. For experimentation we generated train, validation and test splits in a way that no HS-CN pairs occurred across them.
52
+
53
+ <table style="width:25%">
54
+ <tr>
55
+ <th>CONAN-EUS Splits</th>
56
+ <th>Total HS-CN Count</th>
57
+ <tr>
58
+ <td>train</td>
59
+ <td>4833</td>
60
+ </tr>
61
+ <tr>
62
+ <td>validation</td>
63
+ <td>537</td>
64
+ </tr>
65
+ <tr>
66
+ <td>test</td>
67
+ <td>1278</td>
68
+ </tr>
69
+ </table>
70
+
71
+ - 📖 Paper: [Basque and Spanish Counter Narrative Generation: Data Creation and Evaluation]() In LREC-COLING 2024.
72
+ - 💻 Github Repo (Data and Code): [https://github.com/ixa-ehu/conan-e/](https://github.com/ixa-ehu/conan-e/)
73
+
74
+ The CONAN (COunter NArratives through Nichesourcing) dataset was published by [Chung et al., 2019](https://aclanthology.org/P19-1271.pdf)
75
+ and is publicly available in [https://github.com/marcoguerini/CONAN](https://github.com/marcoguerini/CONAN).
76
+
77
+ ## Data
78
+
79
+ This repository contains the Basque and Spanish CONAN-EUS data, in Machine Translated (MT) and post-edited forms. Furthermore, to facilitate experimentation we also share
80
+ the generated splits of the original CONAN English data ([https://github.com/marcoguerini/CONAN](https://github.com/marcoguerini/CONAN)).
81
+
82
+ + data/eu: train/val/test sets with both the MT post-edited Basque data
83
+ + data/es: train/val/test sets with both the MT post edited Spanish data
84
+ + data/en: train/val/test sets of the original CONAN English data
85
 
86
  ## HS-CN example
87
 
88
  The CONAN dataset includes HS-CN pairs, regarding the topic of Muslims and Islam.
89
+ An example of a HS-CN pair in Basque, Spanish and English is illustrated below:
90
 
91
  | HS | CN |
92
  |-------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------|
93
+ | Musulmanok ez dugu gure kultura aberastu dezakeen ezer erabilgarririk. | Zer egin dute guretzat musulmanek? Beno, kafea, erlojuak, kamerak, higienea, unibertsitateak, tresna kirurgikoak, mapak, musika, aljebra. |
94
+ | Los musulmanes no tienen nada útil que pueda enriquecer nuestra cultura.| ¿Qué han hecho los musulmanes por nosotros? Bueno, café, relojes, cámaras, higiene, universidades, instrumentos quirúrgicos, mapas, música, álgebra.|
95
+ | Muslims do not have anything useful that can enrich our culture.| What have Muslims ever done for us? Well, Coffee, Clocks, Cameras, Hygiene, Universities, Surgical Instruments, Maps, Music, Algebra.|
96
 
97
+ If you use CONAN-EUS please **cite the following paper**:
98
 
99
  ## Citation
100
 
 
106
  publisher = "Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING)",
107
  url = ""
108
  }
109
+ ```
110
+
111
+ If you also use the English splits then you **should also cite the original CONAN paper**:
112
+
113
+ ```bibtex
114
+ @inproceedings{chung-etal-2019-conan,
115
+ title = "{CONAN} - {CO}unter {NA}rratives through Nichesourcing: a Multilingual Dataset of Responses to Fight Online Hate Speech",
116
+ author = "Chung, Yi-Ling and
117
+ Kuzmenko, Elizaveta and
118
+ Tekiroglu, Serra Sinem and
119
+ Guerini, Marco",
120
+ booktitle = "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics",
121
+ year = "2019",
122
+ pages = "2819--2829"
123
+ }
124
+ ```
125
+
126
+ **Contact**: [Rodrigo Agerri](https://ragerri.github.io/)
127
+ HiTZ Center - Ixa, University of the Basque Country UPV/EHU