PaTAT

Overview

• To build and run from source-code
  • Frontend
  • Backend
• To run in docker
  • Backend
  • Frontend
• To run in docker-compose
• To view API docs

Notes:

Currently uses Python 3.9

To build and run from source-code

In two different terminals, do:

Frontend

• cd into frontend
• npm install --legacy-peer-deps
• npm start

Backend

• cd into backend
• create a new environment (for example, using miniconda or python3's venv at the command line, i.e., python3 -m venv ./venv and then venv/bin/activate)
• pip install -r requirements.txt
• python -m spacy download en_core_web_sm
• uvicorn main:app --port 8080 --host 0.0.0.0

To run in docker

In two different terminals, do:

Backend

• cd into backend
• docker build -t patat-backend .
• docker run -p 8080:8080 patat-backend

Frontend

• cd into frontend
• docker build -t patat-frontend .
• docker run -p 3000:3000 patat-frontend

To run in docker-compose

• install docker-compose if you don't have it installed
• docker-compose up --build or docker compose up --build (without the hyphen); seems to be machine dependent. Try one if the other doesn't work!
• if you're running on a local machine, go to http://localhost:3000/; if you're running on an AWS machine (see below) then replace localhost with the public IPv4 address

To view API docs

• head to http://localhost:8080/docs after running the backend

To run in AWS instance

• create an instance with GPU: Instance should have at least 30GBs of local storage
• download the generated keypair and chmod 400 <FILE.pem> to make it read-only
• launch instance if its never been launched; use instance state drop-down to restart a stopped instance if its already been launched by stopped.
• copy the files into the instance (either scp the directory from your local machine scp -i "FILE.pem" -r <DIR> ec2-user@INSTANCE_ADDRESS: or git clone after sshing into the instance ssh -i "FILE.pem" ec2-user@INSTANCE_ADDRESS) where INSTANCE_ADDRESS is the public ipv4 DNS address.
• install docker and docker compose: helpful resources ( to install docker link , to install docker-compose link, to start docker demeaon link
• Using the Instance UI in Amazon AWS, allow ports 3000 and 8000 to be accessible from your instance inbound and outbound security rules
• ssh in and go to PaTAT-pattern-based-thematic-annotation-tool/frontend/src/assets/base_url.jsx and replace the ip address with your instances public ipv4
• follow the running instructions above to run PaTAT

Reference

If you use our tools/code for your work, please cite the following paper:

Simret Araya Gebreegziabher, Zheng Zhang, Xiaohang Tang, Yihao Meng, Elena L. Glassman, and Toby Jia-Jun Li. 2023. PaTAT: Human-AI Collaborative Qualitative Coding with Explainable Interactive Rule Synthesis. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI '23). https://doi.org/10.1145/3544548.3581352