Title
Exploiting taxonomical knowledge to compute semantic similarity: an evaluation in the biomedical domain
Abstract
Determining the semantic similarity between concept pairs is an important task in many language related problems. In the biomedical field, several approaches to assess the semantic similarity between concepts by exploiting the knowledge provided by a domain ontology have been proposed. In this paper, some of those approaches are studied, exploiting the taxonomical structure of a biomedical ontology (SNOMEDCT). Then, a new measure is presented based on computing the amount of overlapping and non-overlapping taxonomical knowledge between concept pairs. The performance of our proposal is compared against related ones using a set of standard benchmarks of manually ranked terms. The correlation between the results obtained by the computerized approaches and the manual ranking shows that our proposal clearly outperforms previous works.
Year
DOI
Venue
2010
10.1007/978-3-642-13022-9_28
IEA/AIE (1)
Keywords
Field
DocType
biomedical domain,semantic similarity,important task,domain ontology,biomedical ontology,biomedical field,concept pair,manual ranking shows,taxonomical knowledge,computerized approach,taxonomical structure,data mining,computational semantics
Ontology (information science),Semantic similarity,Ontology,Information retrieval,Ranking,Computer science,Biomedicine,Semantic computing
Conference
Volume
ISSN
ISBN
6096
0302-9743
3-642-13021-6
Citations 
PageRank 
References 
4
0.50
16
Authors
4
Name
Order
Citations
PageRank
Montserrat Batet189937.20
David Sanchez2637.62
Aida Valls356120.52
Karina Gibert428134.01