HMI Project: Course search

Dependencies:

• Installed docker engine (docker and docker compose must be available via command line): link
• Python

Create vectorised data/sentences (done, only required to recreate data):

1. Create virtual python env
2. Install llm/hss_server/requirements.txt
3. Run sbert_train.py in train

Extract data from PDF

1. Build PDFExtract container by running docker build . -t pdfextract:latest in PDFExtract directory
2. Run the created container: docker run -p 8080:8080 pdfextract
3. Go to localhost:8080/modules_sql to fetch the SQL query required to create the database or to localhost:8080/modules_json to get the JSON representation of the modules
4. (SQLite) Use an SQLiteBrowser to create the table and populate it with data by running the code generated by the extracter.

Running the application

Open a command line and type in:

docker compose -f composer.yml

More detailed documentation can be found in the documentation subdirectory.