Information Retrieval (CS60092)

This repository contains code, data, and write-ups for the Information Retrieval course assignments and resources. It was organized to store three main assignments, datasets, results, and supporting materials (books, papers, slides).

Course

• Course name: Information Retrieval
• Author / Owner: Tuhin Mondal (22CS10087)

Repository Structure

Top-level layout (folders and representative files):

A1-Boolean AND Retrieval Using Inverted Index/
├─ bool.py
├─ indexer.py
├─ parser.py
├─ README.md
├─ requirements.txt
├─ run.sh
├─ Dataset/
└─ Output/
  ├─ queries.txt
  └─ results.txt

A2-Scoring and Evaluation/
├─ ranker.py
├─ evaluator.py
├─ README.md
├─ requirements.txt
├─ run.sh
├─ Dataset/
│  ├─ TIME_Documents.txt
│  ├─ TIME_Queries.txt
│  └─ TIME_Relevance.txt
├─ Evals/
└─ Ranks/

A3-Wordnet based Summarization/
├─ summarizer.py
├─ evaluator.py
├─ README.txt
└─ BBC_News_1K/
  ├─ Generated Summaries/
  ├─ News Articles/
  ├─ Summaries/
  └─ Evals/

Books/
Research Papers/
Slides/

Dependencies

• Each assignment maintains its own requirements.txt. Common dependencies for the course include NLTK, NumPy, and scikit-learn (check each requirements.txt for exact pins).

Data and Outputs

• Dataset/ directories inside assignments contain the raw documents and query/relevance files used for experiments.
• Output/, Evals/, and Ranks/ folders contain generated outputs, ranked lists, and evaluation metrics. Do not overwrite these if you want to preserve results.

Course Details / Purpose

• The repository is structured to support learning and evaluation of Information Retrieval concepts:
  • Boolean retrieval via inverted indexes
  • Scoring and ranking with evaluation against ground truth relevance
  • Summarization methods using lexical resources (WordNet)