Ukwabelana: an open-source morphological Zulu corpus

Zulu is an indigenous language of South Africa, and one of the eleven official languages of that country. It is spoken by about 11 million speakers. Although it is similar in size to some Western languages, e.g. Swedish, it is considerably under-resourced. This paper presents a new open-source morphological corpus for Zulu named Ukwabelana corpus. We describe the agglutinating morphology of Zulu with its multiple prefixation and suffixation, and also introduce our labeling scheme. Further, the annotation process is described and all single resources are explained. These comprise a list of 10,000 labeled and 100,000 unlabeled word types, 3,000 part-of-speech (POS) tagged and 30,000 raw sentences as well as a morphological Zulu grammar, and a parsing algorithm which hypothesizes possible word roots and enumerates parses that conform to the Zulu grammar. We also provide a POS tagger which assigns the grammatical category to a morphologically analyzed word type. As it is hoped that the corpus and all resources will be of benefit to any person doing research on Zulu or on computer-aided analysis of languages, they will be made available in the public domain from http://www.cs.bris.ac.uk/Research/MachineLearning/Morphology/Resources/.