{
  "fetched_at": "2026-10-22T09:34:20Z",
  "source_url": "https://www.kaggle.com/competitions/gendered-abuse-detection-shared-task/data",
  "pages": [
    {
      "name": "data-description",
      "content": "We provide a novel dataset on gendered abuse in Hindi, Tamil and Indian English. his dataset was annotated by eighteen activists and researchers who have faced or studied gendered abuse. The dataset contains a total of 7638 posts in English, 7714 posts in Hindi, and 7914 posts in Tamil. We will provide a train and a test split for the dataset. \n\n## <u>Dataset</u>\nThe dataset has been annotated by asking 3 different questions(labels) to annotators. The dataset contains posts tagged with the following three labels:\n1. **Label 1** - Is this post gendered abuse when not directed at a person of marginalized gender and sexuality?\n2. **Label 2** - Is the post gendered abuse when directed at a person of marginalized gender and sexuality?\n3. **Label 3** - Is this post explicit/aggressive?\n\n### The values for each label could be the following \n1. \"1\" - Indicates the annotator believes the post (tweet) matches the label.\n2. \"0\" - Indicates the annotator doesn’t believes the post (tweet) does not match the label.\n3. \"NL\" - The post was assigned to the annotator but not annotated.\n4. \"NaN\" - Indicates the post was not assigned to the annotator\n\n## <u>Files</u>\n### 1) `train.csv`\nThis is the **train data** file for the completion. For simplicity we have clubbed the entire dataset into one file i.e. annotation for all the languages and labels are merged into one file. \nEach tweet has an annotation for all the three labels, so in this dataset, you will find each tweet present three times. \n**Column's Explained**\n- `id` - a sequential number given to a row\n- `text` - the text of the tweet\n- `language` - the language of the tweet\n- `key` - the label for which the tweet is annotated. `question_1` is **label 1**, `question_2` is **label 2** and `question_3` is **label 3**\n- `en_a1` ... `en_a6` - columns for English annotators, these columns present the values for each tweet annotated by the annotators.\n- `hi_a1` ... `hi_a5` - columns for Hindi annotators, these columns present the values for each tweet annotated by the annotators.\n- `ta_a1` ... `ta_a6` - columns for Tamil annotators, these columns present the values for each tweet annotated by the annotators.\n\n> [NOTE]  \n> The English annotators have just annotated the English tweet, the Hindi annotators have annotated the Hindi tweets and the same for Tamil annotators. \n\n### 2) `test_label1.csv`\nThis is the test file for **label 1**\n**Column's Explained**\n- `id` - a sequential number given to a row\n- `text` - the text of the tweet\n- `language` - the language of the tweet\n- `key` - the label for which the tweet is annotated. All the values in this column for this file will be `question_1`\n\n### 3) `test_label3.csv`\nThis is the test file for **label 3**\n**Column's Explained**\n- `id` - a sequential number given to a row\n- `text` - the text of the tweet\n- `language` - the language of the tweet\n- `key` - the label for which the tweet is annotated. All the values in this column for this file will be `question_3`",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "Online gender-based violence is a growing challenge that compounds existing social and economic vulnerabilities. It can cause people to recede from online spaces impacting their political and economic opportunity. At its worst, it can lead to loss of life. While there is a need for automated approaches to detect gendered abuse, there is a lack of Indic language datasets that enable such approaches for Indian language content.\n\nWe provide a novel dataset on gendered abuse in Hindi, Tamil and Indian English. his dataset was annotated by eighteen activists and researchers who have faced or studied gendered abuse. The dataset contains a total of 7638 posts in English, 7714 posts in Hindi, and 7914 posts in Tamil.\n\nMore information on how the dataset is created and its details, kindly refer to the dataset paper provided below. \n**Link to the Dataset Paper** - https://arxiv.org/abs/2311.09086\n\nThe annotations and each label is being explained in the [**Data**](https://www.kaggle.com/competitions/gendered-abuse-detection-shared-task/data) section of the competition. ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "Submissions are evaluated using the **F-1 Score**. It presents a good balance between precision and recall and gives good results on imbalanced classification problems.\n\n## Submission File\nFor each ID in the test set, you must predict a probability for the VALUE variable. The file should contain a header and have the following format:\n\n| **id** | **value** |\n|:------:|:---------:|\n|    2   |     1     |\n|    5   |     0     |\n|    6   |     1     |",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "### Team Size\nThe maximum team size is 6.\n\n### Submission Limits\n- You may submit a maximum of 5 entries per day.\n- You may select up to 2 final submissions for judging.\n\n### Competition Timeline\nTraining Set Release - 15th November 2023\nTest Set Release - 6th December 2023\nSubmissions Due - 8th December 2023\nResults Declared - 10th December 2023\n\n### Code and Model Sharing\n- Participants are not allowed to share code or models during the competition.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "This is shared task titled - **Gendered Abuse Detection in Indic Languages** as a part of [ICON 23](http://icon2023.unigoa.ac.in/)\n\nLink to the Shared Task webiste - https://sites.google.com/view/icon2023-tattle-sharedtask/home\n\nTo be **eligible** for the shared task and to **officially register**, you need to send an email to **aatman@tattle.co.in**. Please refer the website for detailed participation guidelines. ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}