Building a computational morphological analyser/generator for Zulu using the Xerox finite-state tools

Zulu is a member of the Bantu language family and is spoken by approximately 8 million people in South Africa. In terms of natural language processing it is certainly a lesser-studied language of the world. This language family is characterised by its rich agglutinating morphological structures. In this paper we focus on particularly complex constructions pertaining to Zulu nouns and verbs, discussing both the morphotactics, and the morphophonological alternation rules that apply. Our approach in building the analyser/generator for Zulu, using the Xerox finite-state tools, in particular lexc and xfst (Beesley and Karttunen, 2003) is explained.