The source code for the article "Topic extraction in chatbot applications".

├── 00-cluster-selection.ipynb
├── 01_topic-extraction.ipynb
├── README.md
├── data
│   └── dev-v2.0.json
├── figures
├── requirements.txt
└── src
    ├── misc.py
    └── nlp.py

The illustrative example is the "dev" dataset from the Stanford Question Answering Dataset. The dataset can be downloaded from https://rajpurkar.github.io/SQuAD-explorer/. The .json file should be placed in the data/ folder.

Create a conda environment from requirements.txt:

conda create python=3.10 --name topic-extraction-env
conda activate topic-extraction-env
pip install -r requirements.txt