Ontologies for a Global Language Infrastructure

Given a situation where human language technologies have been maturing considerably and a rapidly growing range of language data resources being now available, together with natural language processing (NLP) tools/systems, a strong need for a global language infrastructure (GLI) is becoming more and more evident, if one wants to ensure re-usability of the resources. A GLI is essentially an open and web-based software platform on which tailored language services can be efficiently composed, disseminated and consumed. An infrastructure of this sort is also expected to facilitate further development of language data resources and NLP functionalities. The aims of this paper are twofold: (1) to discuss necessity of ontologies for a GLI, and (2) to draw a high-level configuration of the ontologies, which are integrated into a comprehensive language service ontology. To these ends, this paper first explores dimensions of GLI, and then draws a triangular view of a language service, from which necessary ontologies are derived. This paper also examines relevant ongoing international standardization efforts such as LAF, MAF, SynAF, DCR and LMF, and discusses how these frameworks are incorporated into our comprehensive language service ontology. The paper concludes in stressing the need for an international collaboration on the development of a standardized language service ontology.