search-engine

A text-based document search engine that retrieves the top 10 most relevant documents from the 1989/1990 LATimes collection. The system is comprised of the following components:

• Index Engine (Creates lexicon and inverted index)
• Metadata Engine (Saves document metadata)
• Ranking Engine (Generates BM25 rankings)
• Snippet Engine (Generates query-biased Snippets)
• Query Engine (Establishes user flow and SERP)

Setup

To run the QueryEngine program, navigate into search-engine/engine/src

• Build the class files with the following commands:
  • javac QueryEngine.java
  • Errors may arise if you are not using Java 16 (latest version)
• Run the QueryEngine program:
  • java QueryEngine {store_path}

To run the IREvaluator, navigate into search-engine/evaluator

• Build the class files with the following commands:
  • javac IREvaluator.java
  • Errors may arise if you are not using Java 16 (latest version)
• Run the IREvaluator program:
  • java IREvaluator {path_to_results_file} {path_to_qrels_file}