Cluster labeling with Linked Data
Journal of Theoretical and Applied Information Technology (2013)
Authors: Martin Dostal, Michal Nykl, Karel Ježek.
In this article, we would like to introduce our approach to cluster labeling with Linked Data. Clustering web pages into semantically related groups promises better performance in searching the Web. Nowadays, only special semantic search engines provide clustering of results. Other engines are doubtful as far as the quality of clusters and moreover a dependable system for labeling these clusters is lacking. Linked Data is a set of principles for publishing structured data in a machine readable way with regards to linking with other Web resources. This enables data from different sources to be connected and queried over the Internet. The information from Linked Data can be used for preliminary estimates of topics covered by a set of documents. Topics are represented as resources from Linked Data and are used for smooth human-readable labeling of clusters.