Full Text Search engine

A full text search engine implemented as part of HUJI's Web Information Retrieval course.

The engine currently supports a specific dataset - Amazon product review data taken from here, using a line-oriented data format (see the .txt files under datasets for an example)

The main classes of this library are:

• webdata.IndexWriter, for constructing the index given a dataset file

• webdata.IndexReader for querying the index

• webdata.ReviewSearch for performing various text search operations

Documentation

• Click here for an explanation and visualization of the index structure, as well as theoretical runtime analysis of index operations.

• Click here for various benchmarks of index construction and querying.

• Click here for an explanation of a custom product ranking function I've implemented for product search.

Most of the classes and methods were also documented, see below on how to create javadocs.

Build instructions

Requires Java 11+ and Maven.

• Type mvn package to compile, test and package this library, and generate docs.The resulting jars will be located at target.

  Documentation can be found at target/apidocs/index.html

  (Skip testing by adding -Dmaven.test.skip=true)