{
  "id": 396575,
  "title": "Similar additional datasets and introductions",
  "url": "/competitions/benetech-making-graphs-accessible/discussion/396575",
  "author_name": "",
  "post_date": "2023-03-22T06:53:48.037195100Z",
  "votes": 20,
  "comment_count": 1,
  "views": 0,
  "content": "<h2>1. CHART-Infographics</h2>\n<p><a href=\"https://chartinfo.github.io/\" target=\"_blank\">https://chartinfo.github.io/</a></p>\n<h2>2. PlotQA:</h2>\n<p>Dataset webseite: <a href=\"https://iitmnlp.github.io/PlotQA/\" target=\"_blank\">https://iitmnlp.github.io/PlotQA/</a><br>\nOriginal paper: <a href=\"https://arxiv.org/pdf/1909.00997.pdf\" target=\"_blank\">https://arxiv.org/pdf/1909.00997.pdf</a></p>\n<p>This article introduces a new dataset called PlotQA, which aims to bridge the gap between existing synthetic datasets for reasoning over plots and real-world plots. The dataset contains 28.9 million question-answer pairs over 224,377 plots sourced from real-world data, with questions based on crowd-sourced question templates. The article also evaluates existing state-of-the-art models on PlotQA and proposes a hybrid approach that can handle fixed vocabulary as well as out-of-vocabulary questions. </p>\n<h2>3. ChartQA</h2>\n<p>Original paper: <a href=\"https://arxiv.org/pdf/2203.10244v1.pdf\" target=\"_blank\">https://arxiv.org/pdf/2203.10244v1.pdf</a></p>\n<p>This article introduces a new benchmark dataset for Chart Question Answering (ChartQA) that focuses on complex reasoning questions involving visual and logical operations. The dataset contains both human-written and automatically generated questions, and real-world charts from various sources. The article also proposes a pipeline approach that combines visual features and extracted data from charts to train transformer-based QA models that achieve state-of-the-art results. The related work section discusses existing datasets, models, and chart data extraction techniques.</p>\n<h2>4. DVQA</h2>\n<p><a href=\"https://github.com/kushalkafle/DVQA_dataset\" target=\"_blank\">https://github.com/kushalkafle/DVQA_dataset</a><br>\nOriginal paper: <a href=\"https://arxiv.org/pdf/1801.08163v2.pdf\" target=\"_blank\">https://arxiv.org/pdf/1801.08163v2.pdf</a></p>",
  "messages": [
    {
      "id": "2191722",
      "postDate": "03/22/2023 06:53:48",
      "content": "<h2>1. CHART-Infographics</h2>\n<p><a href=\"https://chartinfo.github.io/\" target=\"_blank\">https://chartinfo.github.io/</a></p>\n<h2>2. PlotQA:</h2>\n<p>Dataset webseite: <a href=\"https://iitmnlp.github.io/PlotQA/\" target=\"_blank\">https://iitmnlp.github.io/PlotQA/</a><br>\nOriginal paper: <a href=\"https://arxiv.org/pdf/1909.00997.pdf\" target=\"_blank\">https://arxiv.org/pdf/1909.00997.pdf</a></p>\n<p>This article introduces a new dataset called PlotQA, which aims to bridge the gap between existing synthetic datasets for reasoning over plots and real-world plots. The dataset contains 28.9 million question-answer pairs over 224,377 plots sourced from real-world data, with questions based on crowd-sourced question templates. The article also evaluates existing state-of-the-art models on PlotQA and proposes a hybrid approach that can handle fixed vocabulary as well as out-of-vocabulary questions. </p>\n<h2>3. ChartQA</h2>\n<p>Original paper: <a href=\"https://arxiv.org/pdf/2203.10244v1.pdf\" target=\"_blank\">https://arxiv.org/pdf/2203.10244v1.pdf</a></p>\n<p>This article introduces a new benchmark dataset for Chart Question Answering (ChartQA) that focuses on complex reasoning questions involving visual and logical operations. The dataset contains both human-written and automatically generated questions, and real-world charts from various sources. The article also proposes a pipeline approach that combines visual features and extracted data from charts to train transformer-based QA models that achieve state-of-the-art results. The related work section discusses existing datasets, models, and chart data extraction techniques.</p>\n<h2>4. DVQA</h2>\n<p><a href=\"https://github.com/kushalkafle/DVQA_dataset\" target=\"_blank\">https://github.com/kushalkafle/DVQA_dataset</a><br>\nOriginal paper: <a href=\"https://arxiv.org/pdf/1801.08163v2.pdf\" target=\"_blank\">https://arxiv.org/pdf/1801.08163v2.pdf</a></p>",
      "rawMarkdown": "## 1. CHART-Infographics\nhttps://chartinfo.github.io/\n\n## 2. PlotQA:\nDataset webseite: https://iitmnlp.github.io/PlotQA/\nOriginal paper: https://arxiv.org/pdf/1909.00997.pdf\n\nThis article introduces a new dataset called PlotQA, which aims to bridge the gap between existing synthetic datasets for reasoning over plots and real-world plots. The dataset contains 28.9 million question-answer pairs over 224,377 plots sourced from real-world data, with questions based on crowd-sourced question templates. The article also evaluates existing state-of-the-art models on PlotQA and proposes a hybrid approach that can handle fixed vocabulary as well as out-of-vocabulary questions. \n\n\n## 3. ChartQA\nOriginal paper: https://arxiv.org/pdf/2203.10244v1.pdf\n\nThis article introduces a new benchmark dataset for Chart Question Answering (ChartQA) that focuses on complex reasoning questions involving visual and logical operations. The dataset contains both human-written and automatically generated questions, and real-world charts from various sources. The article also proposes a pipeline approach that combines visual features and extracted data from charts to train transformer-based QA models that achieve state-of-the-art results. The related work section discusses existing datasets, models, and chart data extraction techniques.\n\n\n## 4. DVQA\nhttps://github.com/kushalkafle/DVQA_dataset\nOriginal paper: https://arxiv.org/pdf/1801.08163v2.pdf",
      "votes": null
    },
    {
      "id": "2199651",
      "postDate": "03/27/2023 22:50:19",
      "content": "<p>Great ! Thanks for sharing, this will be very helpful.</p>",
      "rawMarkdown": "Great ! Thanks for sharing, this will be very helpful.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2199651,
      "author_name": "cid007",
      "author_url": "",
      "post_date": "03/27/2023 22:50:19",
      "content": "<p>Great ! Thanks for sharing, this will be very helpful.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2191722": "## 1. CHART-Infographics\nhttps://chartinfo.github.io/\n\n## 2. PlotQA:\nDataset webseite: https://iitmnlp.github.io/PlotQA/\nOriginal paper: https://arxiv.org/pdf/1909.00997.pdf\n\nThis article introduces a new dataset called PlotQA, which aims to bridge the gap between existing synthetic datasets for reasoning over plots and real-world plots. The dataset contains 28.9 million question-answer pairs over 224,377 plots sourced from real-world data, with questions based on crowd-sourced question templates. The article also evaluates existing state-of-the-art models on PlotQA and proposes a hybrid approach that can handle fixed vocabulary as well as out-of-vocabulary questions. \n\n\n## 3. ChartQA\nOriginal paper: https://arxiv.org/pdf/2203.10244v1.pdf\n\nThis article introduces a new benchmark dataset for Chart Question Answering (ChartQA) that focuses on complex reasoning questions involving visual and logical operations. The dataset contains both human-written and automatically generated questions, and real-world charts from various sources. The article also proposes a pipeline approach that combines visual features and extracted data from charts to train transformer-based QA models that achieve state-of-the-art results. The related work section discusses existing datasets, models, and chart data extraction techniques.\n\n\n## 4. DVQA\nhttps://github.com/kushalkafle/DVQA_dataset\nOriginal paper: https://arxiv.org/pdf/1801.08163v2.pdf",
    "2199651": "Great ! Thanks for sharing, this will be very helpful."
  },
  "source": "meta"
}