Using simple ontologies to build personal Webs of knowledge

Abstract

Current AI and Semantic Web research is striving towards solving the problem of retrieving precise knowledge from the Knowledge Web. Still, thousands of retrieved documents (most of them irrelevant) are flooding Web users who are unable to control and validate them against their queries. Although structuring Web documents is currently the most acceptable solution to the problem, it seems that the problems of a) "thousands of documents returned for a query" and b) neglecting the unstructured repositories, still remain. In this paper, we present our approach to these problems, which does not necessarily require the transformation of the current Web. We show how structured and unstructured documents could be conjunctively queried using simple ontologies built-up by the queries. More importantly, we show how personal collections of Web documents can be built and queried, so that retrieval within such structures can result in precise knowledge. © 2006 Springer-Verlag London.