Information-Retrieval-System

• Introduction
• PRD: Information Retrieval System
• API Documentation
  • Search
• How to run and Install
• Technologies and Tools

Introduction

This is a simple information retrieval system that can be used to search for documents in a corpus. The system is based on the positional index data structure. The system is built using Node.js and Express.js. The system is built as a part of the course CS F313 Data Storage and Retrieval at Helwan University Software Engineering program 2023/2024.

PRD: Information Retrieval System

1. ✅ Tokenization should be done using the following rules:

   • ✅ Splitting on whitespace.
   • ✅ Removing punctuation.
   • ✅ Removing stop words.
   • ✅ Stemming using the Porter Stemmer algorithm.

2. ✅ Constructing Auxiliary structure(s) (Positional index) to speed up the search process should be done.

3. ✅ Phrase query search should be supported using the positional index.

4. ✅ Term frequency and inverse document frequency should be used to rank the documents in the corpus.

5. ✅ Idf smoothing should be used to rank the documents.

6. ✅ TF.idf matrix should be normalized using the cosine normalization technique.

7. ✅ Similarity between query and each document should be computed using the cosine similarity measure.

8. ✅ Boolean query search should be supported using the positional index.

API Documentation

Search

`GET` /search

body: {
    query: string
}

How to run and Install

# Clone the repo
git clone https://github.com/Adosh74/Information-Retrieval-System

# Install dependencies
yarn install
or
npm install

# Run the server
yarn start

# Run the server in development mode
yarn start:dev

• After running the server, you can access the API through the following URL: http://localhost:3001

• And display all PRD tables in console

Technologies and Tools

• Node.js
• Express.js
• Typescript