centre-for-humanities-computing / classic-topic

Web application for dynamic topic modelling on classical texts in the "Computing antiquity" project

Topiquity

Web application for dynamic topic modelling on classical texts in the "Computing antiquity" project.

Setup

In order for the application to run properly you will need a dat/cleaned_corpus.csv in the working directory.
This file has to contain a table with one column being id_nummer containing the IDs of the texts, the other being text containing a whitespace-separated cleaned texts.

Debugging

For debugging purposes (or if you're just lazy to mess with docker) you may run the app by installing dependencies:

pip install -r requirements.txt

Then run the main script in the following fashion:

cd src && python3 main.py

The server will listen at localhost:8080, you are free to open it up in your browser.

Deployment

Both a Dockerfile and a docker compose default configuration is included in the repository, these might aid you in deploying the application to a webserver of some sorts, or just run the app on your computer in an isolated environment. It's easiest to get started by running the app in docker compose:

docker compose up

This will spin up a container and attach to localhost:7011

Known limitations

• DMM and LSI don't work particularly well, sometimes not at all, you should use LDA or NMF as things stand right now. A project-agnostic topic investigation platform might be developed in the near future.
• The code is heavily project-specific, if you want to use the application with a different corpus you will have to dig deep into the code.