UMass at TREC WEB 2014: Entity Query Feature Expansion using Knowledge Base Links

Abstract

Entity linking tools predict links between entity mentions in text and knowledge base entries. In this work we leverage the rich semantic knowledge available through these links to understand relevance of documents for a query. We focus on the ad hoc task on the category A subset and demonstrate the benefit of entity-centric approaches even for non-entity queries like "dark chocolate health benefits".

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 2014
Accession Number
ADA618672

Entities

People

  • Laura Dietz
  • Patrick Verga

Organizations

  • University of Massachusetts Amherst

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Abstracts
  • Agriculture
  • Automatic
  • Chocolate
  • Data Sets
  • Electronic Mail
  • Feedback
  • Indicators
  • Information Operations
  • Information Retrieval
  • Language
  • Plants
  • Standards
  • Test Sets
  • Training
  • Trees
  • United States

Fields of Study

  • Computer science

Readers

  • Computational Linguistics
  • Information Retrieval
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Information Retrieval