Conjunctive Inverted Index

This repository contains two minimal, well-tested implementations for multi-term search over documents:

• src/inverted_index.py: a classic inverted index with set intersections for AND queries.
• src/conjunctive_index_binary.py: a binary-encoded variant that treats each document’s term set as a bitmask and answers conjunctive (AND) queries via bitwise checks.

The goal is to show simple, clear baselines for conjunctive (multi-term) search and highlight trade-offs between straightforward posting-list intersections and compact bitmask-based filtering.

Article

For background and motivation, see the article:

• From Inverted Index to Conjunctive Inverted Index: Optimizing Multi-Term Queries

Installation

Requirements:

• Python 3.8+

Install dev dependency for tests:

pip install -r requirements.txt

Tests

This repo is test-first. Run the full test suite with:

pytest -q

Relevant tests:

• tests/test_inverted_index.py
• tests/test_conjunctive_index_binary.py