Exploiting Citation Networks in Large Corpora to Improve Relevance on Broad Queries

Explore an innovative algorithm for improving search relevance in large legal databases through citation network analysis. Learn how to rank documents by authority for broad queries in unstructured text corpora containing millions of lengthy documents. Discover the mathematical foundations, Lucene/Solr implementation, and production deployment of this approach that merges traditional ranking methods with citation-based authority ranking. Gain insights into rescuing relevant documents initially missed due to lack of query terms and understand how this technique meets user expectations for intuitive search results in legal information systems.

Haystack US 2023 - Marc André Morissette: Exploiting Citation Networks in Large Corpora