SLS : Neural Information Retrieval(IR)-based Semantic Search model
An autonomous research agent that synthesizes detailed reports by combining dynamic planning, unsupervised topic discovery (PCA/KMeans) HyDe (hypothetical doc embeddings), and a self-correcting reflexion loop.
A fully interactive domain-specific chatbot implemented using Prolog and PySwip.
About Twitter Searcher is a search engine which searches documents in the corpus of tweets scrapped from your Twitter homepage. The search engine is the implementation of the Vector Space Model. The tweets are scrapped using selenium and stored in MongoDB as a corpus. MongoDB is a document-based Database System.
Context that works
Solutions from the Advent of Haystack 2024 event, exploring Haystack framework fundamentals through advanced RAG pipelines and intelligent agents
Project repository for the development of a Question-Answering (QA) information retrieval system fine-tuned on customer queries.
A Persian Information Retrieval system implementing text preprocessing, inverted index, TF-IDF ranking, and document retrieval using the Hamshahri news corpus.
A complete IR system.
An information retrieval system for boolean queries, proximity quries and wildcard queries using Inverted indexing, Biword indexing, positional indexing and soundex indexing.
Moogle!, A basic System of Information Retrieval. First Programming Project
SpeciFic is an NLP fanfiction recommender for AO3 that compares Knowledge Graph and semantic-embedding (FAISS) retrieval approaches in an automated evaluation framework.
Complex Query Synthesis for Enhanced Information Retrieval
Library Management System is an application which refers to library systems which are generally small or medium in size. It is used by librarian to manage the library using a computerized system where he/she can add new books and Page sources.
Production-oriented Elasticsearch 8.15.3 search platform built with Django REST Framework and Clean Architecture. Index design, text analysis, relevance, fuzzy search, autocomplete, facets, filtering, pagination, and explainability — exposed over a documented HTTP API.
Treats a retrieval pipeline (BM25 + dense + fusion + rerank) as a search space: diagnoses why queries fail, proposes targeted config mutations, and reports gains on a held-out split with significance testing. BM25 and the search loop built from scratch.
Rank based information retrieval system. Ranking done based on Tf-Idf scores of documents and queries
Multi-turn conversational shopping agent over a 50,000-product catalog. BM25 + SQLite FTS5, zero dependencies.
Retrieval over chemistry papers keyed on structure, not on the name someone happened to type. Names resolve to InChIKey skeletons at ingest.
To associate your repository with the information-retrieval-system topic, visit your repo's landing page and select "manage topics."