bm25-okapi

My implementation for a kaggle competition: https://www.kaggle.com/competitions/WattBot2025

Hybrid Retrieval System combining keyword matching (BM25) with semantic similarity (Vectorstore) for improved retrieval.

BM25 (Okapi BM25) implementation in TypeScript with field boosting and parallel processing support.

B-SERS: Books Search Engine and Recommendation System. This system combines the functionalities of a search engine and a recommendation system to effectively address the challenges of finding and suggesting books

Implementation of BM25/Okapi and BIM models for probabilistic document ranking in Information Retrieval.

RAG Application with Contextual Retrieval and Lexical Retrieval.

Prodapt - AI & Data Science Track/Agentic RAG over Mixed Data Sources (Track - C)

Project Search Enghine UAS Temu Kembali Informasi menggunakan dataset Indonesian Hoax News dengan metode BM25

Grameenphone customer service chatbot with LSTM and BM25

A distributed search engine with a C++ Crawler/Indexer, Python Ranker (BM25), and Ruby on Rails Interface. Uses RocksDB & Redis

An Information Retrieval system that processes and ranks news articles. It parses XML files, applies stop-word removal and stemming, and uses TF-IDF and BM25 algorithms to score documents against user queries, sorting them by relevance.

An Okapi BM25 implementation in pure Erlang

Local Agentic Hybrid RAG (FastAPI + Chroma + BM25 + Ollama)

Repository for the CENG543 Course Information Retrieval Term Project

RAG evaluation platform comparing LLMs, retrievers, and prompts with cost, latency, and faithfulness scorecards.

This project is an implementation of graph retrieval concept, where knowledge graphs are used to capture the entities and relationships to make retrieval better. This project advances my older project naming hybrid_rag by adding this graph retrieval.

Chatbot RAG oparty na sekcji "pomoc" z allegro.