QUT ielab at CLEF 2017 e-Health IR Task: Knowledge Base Retrieval for Consumer Health Search
| Main Authors: | Jimmy, Jimmy, Zuccon, Guido, Koopman, Bevan |
|---|---|
| Format: | Proceeding PeerReviewed application/pdf |
| Terbitan: |
, 2017
|
| Subjects: | |
| Online Access: |
http://repository.ubaya.ac.id/37161/1/Jimmy_2017_QUT%20IELab.pdf http://repository.ubaya.ac.id/37161/ |
Daftar Isi:
- In this paper we describe our participation to the CLEF 2017 e-Health IR Task [6]. This track aims to evaluate and advance search technologies aimed at supporting consumers to and health advice online. Our solution addressed this challenge by developing a knowledge base (KB) query expansion method. We found that the two best KB query expansion methods are mapping entity mentions to KB entities by performing exact matching entity mentions to the KB aliases (EM-Aliases) and multi-matching entity mentions to all KB features (Title, Categories, Links, Aliases, and Body) (EM-All). After mapping between entity mentions to KB entities established, we found the Title of the mapped KB entities as the best source of expansion terms compared to the aliases or combination of both features. Finally, we also found that Relevance Feedback and Pseudo Relevance Feedback are effective to further improve the query effectiveness.