This README file provides an overview of the Medical Information Retrieval System developed for Phase A/Phase B of the assignment. That we completed in the third year of our studies (Computer Science University of Crete)
This project aims to develop a Medical Information Retrieval System for clinical decision support. The assignment consists of two phases: Phase A and Phase B. Phase A involves system development, while Phase B focuses on system evaluation using predefined datasets and metrics.
In Phase A, we have implemented a Medical Information Retrieval System. The system features the following components:
- Data Ingestion: The system ingests a collection of medical documents, including metadata, abstracts, and full texts.
- Indexing: It creates an index for efficient retrieval, enabling users to search for medical information.
- Query Processing: The system processes user queries, identifies relevant documents, and ranks them based on relevance.
- Retrieval Interface: An interface allows users to input queries and retrieve relevant medical documents.
In Phase B, the system developed in Phase A is evaluated for its effectiveness in retrieving medical information. The evaluation process includes the following steps:
- Data Collection: We use the TREC 2014 collection, which includes medical topics, documents, and relevance judgments.
- Automated Evaluation: We have developed an automated evaluation system that calculates several metrics, including bpref, AveP', and nDCG', to assess the retrieval effectiveness.
- Reporting: The evaluation results are reported, including an analysis of which medical topics performed well and areas for improvement.
The results obtained from the Phase B evaluation are available in folder \FinalCollectionIndex
This project is licensed under the MIT License