Towards a bilingual SMS parser for HIV and AIDS information retrieval in Botswana

Abstract

Integrated Health Information System through Mobile Telephony (IHISM) is a project carried out by Computer Science department at the University of Botswana. The main objective of the research project is to develop an integrated healthcare information service which will provide access to a variety of HIV AND AIDS related information for different users. The aim of this paper is to analyze the use of SMSs as an information access technology in a multilingual environment such as Botswana. The main focus is the analysis of the users' language in constructing SMSs to retrieve information from the IHISM information service. The results acquired in this work give a good lead towards designing the most appropriate bilingual natural language parser. The analysis, carried out on a sample of 1,600 SMS queries (800 English, 800 Setswana), would support the way parsing techniques could best hypothesize as accurately as possible the SMS question being asked.