Search
 
loader loading subjects...
cover
Document Similarity
Marie Duží, Marek Menšík, Michal Perdek
241 - 254
10.3233/978-1-61499-177-9-241
Frontiers in Artificial Intelligence and Applications
Volume 251: Information Modelling and Knowledge Bases XXIV
Abstract

This paper deals with document similarity based on contextual similarity of words occurring in the documents. The method makes it possible to cluster documents in which words with similar meanings occur even though the words and their meanings are not identical. These similarities can be discovered due to context tracing. On the other hand, we can distinguish between homonyms bearing different meanings. Thus the proposed method provides a fine-grained information mining from particular documents.

$35.00 / €27.50 / £22.00
Add PDF to cart