Towards Automatic Learning of a Structure Ontology For Technical Articles

Abstract

Despite the high level of success attained by keyword based information retrieval methods, a significant fraction of information retrieval tasks still needs to take into account the semantics of the data. We propose a method that combines an hand-crafted ontology with a robust inductive inference method to assign semantic labels to pieces of technical articles available on the Web. This approach, together with a query language developed for the purpose, supports queries that cannot be resolved using currently available tools. We present preliminary results that describe the precision of the assigned labels and the accuracy of the replies to the semantic queries present to the system.