Ontosophie: A Semi-Automatic System for Ontology Population from Text

This paper describes a system for semi-automatic population of ontologies with instances from unstructured text. It is based on supervised learning, learns extraction rules from annotated text and then applies those rules on new articles for ontology population. Therefore, the system classifies stories and populates a hand-crafted ontology with new instances of classes defined in it. It is based on three components: Marmot - a natural language processor; Crystal - a dictionary induction tool; and Badger - an information extraction tool. A part of the entire cycle is a user who accepts, rejects or modifies newly extracted and suggested instances to be populated. A description of experiments performed with text corpus consisting of 91 articles is given in turn. The results cover the paper and support a presented hypothesis of assigning a rule confidence value to each extraction rule for improving the performance.