A Bayesian model for entity type disambiguation

  • Authors:

  • Barbara Bazzanella;Heiko Stoermer;Paolo Bouquet

  • Affiliations:

  • University of Trento, Dept. of Cognitive Science, Trento, Italy;Fondazione Bruno Kessler, DKM Unit, Trento, Italy;University of Trento, Dept. of Information Science and Engineering, Trento, Italy

  • Venue:
  • AIMSA'10 Proceedings of the 14th international conference on Artificial intelligence: methodology, systems, and applications
  • Year:
  • 2010

Quantified Score

Hi-index 0.02

Visualization

Abstract

Searching for information about individual entities such as persons, locations, events, is a major activity in Internet search. General purpose search engines are the most commonly used access point to entity-centric information. In this context, keyword queries are the primary means of retrieving information about specific entities. We argue that an important first step to understand the information need that underly an entity-centric query is to understand what type of entity the user is looking for. We call this process Entity Type Disambiguation. In this paper we present a Naive Bayesian Model for entity type disambiguation that explores our assumption that an entity type can be inferred from the attributes a user specifies in a search query. The model has been applied to queries provided by a large sample of participants in an experiment performing an entity search task. The results of the model performance evaluation are presented and discussed. Finally, an extension of the model that includes the distribution of attributes inside the query is presented and its implications are investigated.