Ontology-driven, unsupervised instance population

The Semantic Web’s need for machine understandable content has led researchers to attempt to automatically acquire such content from a number of sources, including the web. To date, such research has focused on “document-driven” systems that individually process a small set of documents, annotating each with respect to a given ontology. This article introduces OntoSyphon, an alternative that strives to more fully leverage existing ontological content while scaling to extract comparatively shallow content from millions of documents. OntoSyphon operates in an “ontology-driven” manner: taking any ontology as input, OntoSyphon uses the ontology to specify web searches that identify possible semantic instances, relations, and taxonomic information. Redundancy in the web, together with information from the ontology, is then used to automatically verify these candidate instances and relations, enabling OntoSyphon to operate in a fully automated, unsupervised manner. A prototype of OntoSyphon is fully implemented and we present experimental results that demonstrate substantial instance population in three domains based on independently constructed ontologies. We show that using the whole web as a corpus for verification yields the best results, but that using a much smaller web corpus can also yield strong performance. In addition, we consider the problem of selecting the best class for each candidate instance that is discovered, and the problem of ranking the final results. For both problems we introduce new solutions and demonstrate that, for both the small and large corpora, they consistently improve upon previously known techniques.