serena-fang / rest-api

Contains source code of the Python+Flask based REST API for VitaLITy tasks e.g., find similar papers.

vitaLITy - REST API

Requirements:

• Python 3.8+
• Conda (similar to virtualenv + pip)

Setup

• Create a new environment: conda create --name env_vitality_rest_api anaconda or conda create --name env_vitality_rest_api python=3.8 (based on specific python version)
• Activate the environment: conda activate env_vitality_rest_api
• Add conda-forge to the channel: conda config --append channels conda-forge
• Install requirements: conda install --yes --file requirements.txt

Run

• Configure options pertaining to the input data source, its type in config.py

• python server.py - wait until you see three logs: "loaded data" and "created query index for glove" and "created query index for specter"; this may take 3-4 minutes and look like the below screenshot:

• The REST-API will then be available at http://localhost:3000. Use this in the frontend/src/components/App.tsx file as the baseUrl variable.

Topic API configuration

• GET /getPapers(): Returns all papers

• GET /getTopics(): Returns all topics

• POST /getRepresentativePapers(): Returns representative papers by Topic ID, Topic Keyword

  {
      "input_data": ["bias", "uncertainty"], // [list<Keyword>], [list<ID>] depending on the input_type below
      "input_type": "Keyword", // Keyword, ID
      "limit": 3, // any number less than the number of papers available.
  }

• POST /getSimilarTopicsByTopic(): Returns similar topics by topics

  {
      "input_data": [3, 4, 5], // [list<Keyword>], [list<ID>] depending on the input_type below
      "input_type": "ID", // Keyword, ID
      "limit": 3, // any number less than the number of papers available.
  }

• POST /getSimilarTopicsByPaper(): Returns similar topics by ID, Title

  {
      "input_data": ["An Intermittent Click Planning Model.", "Tool Extension in Human-Computer Interaction."], // [list<Title>], [list<ID>] depending on the input_type below
  "input_type": "Title", // Title, ID
  "limit": 3, // any number less than the number of papers available.
  }

• POST /getTopicsByKeyword(): Returns similar topics by Keywords

  {
      "input_data": ["VisUaLiSAtion"], // [list<Keywords>], case insensitive
      "limit": 3, // any number, -1 for all
  }

API configuration

• GET /getPapers(): Returns all papers

• POST /getSimilarPapers(): Returns similar papers by ID, Title

  {
      "input_data": ["An Intermittent Click Planning Model.", "Tool Extension in Human-Computer Interaction."], // [list<Title>], [list<ID>] depending on the input_type below
      "input_type": "Title", // Title, ID
      "limit": 3, // any number less than the number of papers available.
      "embedding": "glove", //glove, specter
      "dimensions": "2D" // 2D, nD
  }

• POST /getSimilarPapersByKeyword(): Returns similar papers by Keywords

  {
      "input_data": ["DaTa VisUaLiSAtion"], // [list<Keywords>], case insensitive
      "limit": 3, // any number, -1 for all
  }

• POST /getSimilarPapersByAbstract(): Returns similar papers by Abstract and Title

  {
  "input_data": {
      "title": "Any WIP Title", // string
      "abstract": "Any WIP abstract" // string
  },
  "limit": 3
  }

• POST /checkoutPapers(): Download a list of dictionaries (JSON) of papers by ID, Title.

  {
      "input_data": ["An Intermittent Click Planning Model.", "Tool Extension in Human-Computer Interaction."], // [list<Title>], [list<ID>] depending on the input_type below
      "input_type": "Title" // Title, ID
  }

Credits

vitaLITy was created by Arpit Narechania, Alireza Karduni, Ryan Wesslen, and Emily Wall.

Citation

@article{narechania2021vitality,
  title={vitaLITy: Promoting Serendipitous Discovery of Academic Literature with Transformers \& Visual Analytics},
  author={Narechania, Arpit and Karduni, Alireza and Wesslen, Ryan and Wall, Emily},
  journal={IEEE Transactions on Visualization and Computer Graphics},
  year={2021},
  doi={10.1109/TVCG.2021.3114820},
  publisher={IEEE}
}

License

The software is available under the MIT License.

Contact

If you have any questions, feel free to open an issue or contact Arpit Narechania.