Tapping the Neglected Resource in Kiswahili Terminology: Automatic Compilation of the Domain-specific Terms from Corpus

Abstract

This paper is about corpus-based computational terminography. It highlights computational techniques and tools for compiling Kiswahili terms in the domain-particular texts written by field experts. At present, a reasonable number of domain-specific texts are available into Kiswahili. However, no efforts have been made so far to collect terms in such texts and integrate them in Kiswahili terminology. The absence of appropriate techniques and tools for carrying out the task should be the explanation for this lack of interest. But recent individual efforts in the development of computational tools for Kiswahili language, particularly by Arvi Hurskainen, at the University of Helsinki, paves the way for using computers for compiling the terms. This paper focuses on the application of Kiswahili-based computational tools developed at the University of Helsinki for describing and using corpus-based term formation patterns for the automatic compilation of Kiswahili terms. The presentation postulates the central role of the computer lexicon in carrying out such a task. The computer lexicon plays a central role in annotating a corpus for the description of the term formation patterns and for the compilation of terms from the text. The paper provides a test case to verify the applicability of the techniques and the tools in the actual compilation of the terms.