Skip to content

Repository for the COLING 2022 paper on Medical Question Understanding and Answering.

Notifications You must be signed in to change notification settings

KhalilMrini/Medical-Question-Answering

Repository files navigation

Medical Question Understanding

1. Paper

This repository contains the code for the following paper:

  • "Medical Question Understanding and Answering with Knowledge Grounding and Semantic Self-Supervision"
    Khalil Mrini, Harpreet Singh, Franck Dernoncourt, Seunghyun Yoon, Trung Bui, Walter Chang, Emilia Farcas, Ndapa Nakashole
    COLING 2022
    Gyeongju, Republic of Korea, and Online
  • 2. Installation

    Use the following commands to install the requirements for this repository:

    pip install -r requirements.txt
    

    3. Training Commands

    The main.py file contains all the training and testing modes.

    For example, use python3 main.py train_meqsum to train a model for the MeQSum dataset.

    4. Chatbot

    You may train a summarizer for the GAR baseline by running the train_summarizer_for_gar.py file.

    Run the app_chatbot.py file to start the Chatbot and evaluate answers. The DPR and GAR take about 40 to 50 seconds to retrieve an answer, but our model should be near immediate to retrieve one.

    About

    Repository for the COLING 2022 paper on Medical Question Understanding and Answering.

    Resources

    Stars

    Watchers

    Forks

    Releases

    No releases published

    Packages

    No packages published