AfriBooms: An Online Treebank for Afrikaans

AfriBooms: An Online Treebank for Afrikaans Liesbeth Augustinus∗, Peter Dirix∗, Daniel van Niekerk†, Ineke Schuurman∗, Vincent Vandeghinste∗, Frank Van Eynde∗, Gerhard van Huyssteen† ∗Centre for Computational Linguistics KU Leuven (University of Leuven), Leuven, Belgium {firstname.lastname}@kuleuven.be †Centre for Text Technology North-West University, Potchefstroom, South Africa {firstname.lastname}@nwu.ac.za Abstract Compared to well-resourced languages such as English and Dutch, natural language processing (NLP) tools for Afrikaans are still not abundant. In the context of the AfriBooms project, KU Leuven and the North-West University collaborated to develop a first, small treebank, a dependency parser, and an easy to use online linguistic search engine for Afrikaans for use by researchers and students in the humanities and social sciences. The search tool is based on a similar development for Dutch, i.e. GrETEL, a user-friendly search engine which allows users to query a treebank by means of a natural language example instead of a formal search instruction. Keywords: Afrikaans, treebank, dependency parser, corpus search tool 1. Introduction Afrikaans is a West Germanic language spoken as a first language by about 7 million people in South Africa and Namibia and by many millions more as a second language. It can be considered a daughter language of Dutch, as it originates in 17th-century Dutch dialects, brought to south- ern Africa by settlers from the Netherlands. Although there are some influences from Malay, Portuguese, Bantu, and Khoisan languages, Dutch and Afrikaans are still more or less mutually comprehensible. One of the main distinctions between Afrikaans and Dutch is a simplification of Dutch morphology in Afrikaans, e.g. dropping the nominal gen- der distinction and only keeping two verb forms for all but the most common verbs (present/infinitive and past partici- ple). In recent years, several NLP tools have been developed for Afrikaans, cf. Grover et al. (2011) for an overview of the available tools. Compared to well-resourced languages such as English and Dutch, however, there are fewer well- performing tools available for Afrikaans. In this paper, we describe the development of an Afrikaans treebank (Section 2.), a dependency parser (Section 3.), and a tool for querying the treebank (Section 4.). 2. Treebank The Afrikaans part of the NCHLT1 Annotated Text Corpora (Puttkammer et al., 2014) was used as a basis for the devel- opment of the treebank. This corpus was selected because it had already been annotated with part-of-speech (POS) tags and word lemmas in a previous project (Eiselen and Puttkammer, 2014). The scope of work described below thus involved additional annotation for word dependencies within sentence context given the pre-existing POS tags and lemmas. 1South African National Centre for Human Language Tech- nologies 2.1. Development The NCHLT text corpus, consisting mainly of government domain documents, contains a number of incomplete sen- tence fragments which could not be annotated sensibly for dependency structure. The first action performed was fil- tering the corpus so that only complete sentences were an- notated. The POS annotation in the NCHLT corpus was based on a fine-grained tag set developed specifically for Afrikaans (Pilon, 2005). This included …

Content truncated to 500 words