↕️ Intuitive axiomatic retrieval experimentation.

Create PyTerrier compatible dense indices using any sentence_transformers model

🔍 Use the ChatNoir search engine in PyTerrier.

Multi-stage Retrieval using SPLADE or RM3 and T5.

Word2vec, sentenceBert, BM25 and IVFFlat Index quality and speed comparison

Exploiting the PyTerrier library to build a Search Engine and resolve the Near Duplicate Detection tasks.

This repository hosts the implementation of a Simple Search Engine designed for efficient information retrieval. The project encompasses several stages from data collection to evaluation, ensuring a comprehensive approach to search and retrieval.

This repository contains the code for a research project that implements and evaluates local word embeddings based on co-authorship and citations for query expansion in PyTerrier on the TREC-Covid dataset.

Fact Finder - a Fact Search Engine

This project creates a basic search engine for text documents, covering data collection, preprocessing, indexing, query processing, expansion, UI development, and performance evaluation. Its goal is to efficiently retrieve relevant information from the document collection.

Information retrieval techniques using Pyterrier

This project constructs an ad-hoc information retrieval system using the 𝐷𝑃𝑅𝑊𝑖𝑘𝑖100 dataset with PyTerrier. NLTK handles query processing, including tokenization and stemming. BM25 ranking is used with enhancing performance through optimizations. The system features a minimalistic tkinter-based user interface for an intuitive experience.