Biomedical Information Retrieval

This is a project related to nlp in biomedical domain. There are search functions, tokenizer, ...

Table of contents

• Code
• Data
• Pretrained models
• Final project

Code

This folder includes functions in information retrieval: skipgram, cbow, tokenization, semantic search, ... with simple Tkinter application To run Tkinter application:

python gui.py

Data

This folder includes data for the above code to train models and evaluate:

• Pubmed_JSON: include JSON format files for PMID documents
• Pubmed_XML: include XML format files for PMID documents
• Tweet_data: includes JSON format file for tweets

Pretrained models

This folder includes pretrained models: skipgram and cbow with different epochs, winsize and training number of documents

Final project

This folder includes the complete search tool with visualization for keywords relations from all documents.

• Each edge represents there are connections between the paragraphs and its citation
• Node size represents how many connections from that node to other nodes