Putting Similarity Assessments into Context: Matching Functions
with the User's Intended Operations
Andrea Rodríguez and
Max Egenhofer Modeling and Using Context, CONTEXT-99, Trento, Italy,
P. Bouquet, L. Serafini, P. Brezillon, and F. Castellani (eds.), Lecture Notes in Artificial Intelligence, Vol. 1688, Springer-Verlag, pp. 310-323, September 1999.
Abstract
This paper presents a practical application of context for the
evaluation of semantic similarity. The work is based on a new model
for the assessment of semantic similarity among entity classes that
satisfies cognitive properties of similarity and integrates
contextual information. The semantic similarity model represents
entity classes by their semantic relations (is-a and part-whole)
and their distinguishing features (parts, functions, and
attributes). Context describes the domain of an application that is
determined by the userÕs intended operations. Contextual
information is specified by a set of tuples over operations
associated with their respective entity-class arguments. Based on
the contextual information, a partial word-sense disambiguation can
be achieved and the relevance of distinguishing features for the
similarity assessment is calculated in terms of the
featuresÕ contribution to the characterization of the
application domain.
PDF
Andrea Rodríguez and
Max Egenhofer