Steve Helmreich & Ron Zacharski,
New Mexico State University
The Role of Ontologies in a Linguistic Knowledge Acquisition Task
In this paper, we present BOAS, a web–based software program used for eliciting linguistic information in sufficient detail to automatically generate a machine translation system. It elicits a variety of lexical, morphological, syntactic, and ecological (letters, symbols, punctuation, etc) data. The system also includes a module for converting this elicited information into a format useful for machine translation (e.g., converting paradigm tables into morphological rules, which are realized in a finite state morphological analyzer). We present the ontology of linguistic elements that is implicit in the BOAS architecture and compare this ontology with the GOLD ontology. In designing Boas we made a simplifying assumption that the ontology describes a way of dividing the knowledge about a particular language into modules that can be dealt with independently of one another. In this paper we present some language phenomena that cannot be handled under this assumption. One such phenomena is the movement of inflectional affixes from their head words to another place in the sentence.
The machine translation system that is automatically produced by BOAS is a transfer system, but underlying this transfer-approach is an interlingual understanding of language use, where formal and structural linguistic categories are mapped into real–world semantic categories. These semantic concepts and categories are represented in the Mikrokosmos Ontology (and therefore also in its larger superset, the Omega ontology– see http://omega.isi.edu/). We describe how this ontology is used within machine translation and compare the upper level of this ontology with the SUMO ontology.
We describe our experiences with using BOAS to gather information about Urdu and Somali. We focus on areas that proved difficult for the system. We examine these areas in the light of the GOLD ontology, to see if these are also difficult to represent in those conceptual categories.
Appendix: Screen Shots of Boas
A View of the Navigation Tree
A Sample Knowledge Elicitation Page