Semi-automated Ontology Creation for High-Level Service Classification

Abstract

In this paper, we propose a LexOnt, a semi-automatic ontology creation tool for high-level service ontology. LexOnt uses the Programmable Web directory of services as its corpus of data along with Internet sources such as Wikipedia in order to generate common terms within a service domain and determine generic properties for classification. Currently, Programmable Web classifies services in a flat categorization where each service is manually classified within a single category. Search is limited to attributes that are not related to the semantics of the service, such as protocol or messaging type. When using an ontology for the service API descriptions, services can be automatically classified and queried for according to their attributes. Additionally, services can then be distributed in an ontology-based service discovery system such as GloServ so that semantic registration and querying of services becomes possible.