README.md

Project-NLP

NLP project for Imperial DS Winter School

Structure

.
├── README.md
├── data
│         └── document_parses
├── pre_data
│         ├── dataset0.pkl
│         ├── filteredOntos.pkl
│         └── finalBPETokenizer15757ForBiomedicine.json
├── output
│         ├── Track-3-2-model
│         ├── Track-4-1
│         ├── Track-4-2
│         └── Track-5
└── nlp_project.ipynb

Install

conda create -n ic_ds_nlp python=3.9
activate ic_ds_nlp
pip install tqdm
pip install tokenizers
pip install gensim
pip install nltk
pip install transformers
pip install torch
pip install joblib
pip install sklearn
pip install matplotlib
pip install mpld3
pip install py2neo

Additional Instructions

Track 2.2

nltk.download('all')

Install NLTK using this cell. Only need to run for one time. If failed, download the NLTK manually on https://www.nltk.org.

Part 5

To run this part, install and configure Neo4j first. The official website of Neo4j: https://neo4j.com

Pre-data

Not committed due to too large data