ABSTRACT: Semantic wikis extend wiki platforms with the ability to represent structured information in a machine-processable way. Ontop of the structured information in the wiki, novel ways to search, browse, and present the wiki content become possible. However, while powerful query languages offer new opportunities for semantic search, users typically are not willing to use the syntax of formal query languages. In this work we present an approach to semantic search that combines the expressiveness and capabilities of structured queries with the simplicity of keyword interfaces: Users articulate their information need using keyword queries, which are translated by the system into structured, conjunctive queries. This translation result may result in multiple possible interpretations of the information need, which can then be selected and further refined by the user via facets. We have implemented this approach to semantic search as an extension to Semantic MediaWiki. The results of a user study in the SMW-based community portal semanticweb.org show the efficiency and effectiveness of the approach.
|