Disambiguation of morphological analysis in Bantu languages

Disambiguation of morphological analysis in Bantu languages Arvi Hurskainen Department of Asian and African Studies Box 13 00014 University of Helsinki Finland Arvi.Hurskainen@helsinki.fi Abstract The paper describes problems in dis- ambiguating the morphological analysis of Bantu languages by using Swahili as a test language. The main factors of ambiguity in this language group can be traced to the noun class structure on one hand and to the bi-directional word-formation on the other. In analyz- ing word-forms, the system applied uti- lizes SWATWOL, a morphological pars- ing program based on two-level formal- ism. Disambiguation is carried out with the latest version (April 1996) of the Constraint Grammar Parser (GGP). Sta- tistics on ambiguity are provided. So- lutions tbr resolving different types of ambiguity are presented and they are demonstrated by examples fi'om corpus text. Finally, statistics on the perfor- mance of the disambiguator are pre- sented. Introduction There are five principal factors in Bantu languages which contribute to ambiguous analysis of word- tbrms. First, nouns are grouped into more than ten marked noun classes. The marking of these classes extends across the noun phrase, whereby the noun governs the choice of markers in de- pendent constituents. Second, verbs inflect stem- initially and mark the subject, object, and rela- tive referent by prefixes, whereby the actual form of each prefix is governed by the noun class of the noun it refers to. In addition, verb derivation also adds to the complexity of verbal morphology. Third, reduplication is a productive phenomenon. Because its accurate description in lexicon is not possible, alternative ways in handling it are dis- cussed. Fourth, the majority of Bantu languages have a tone system, but rarely this is indicated in writing. This adds to morphological ambiguity. Fifth, various semantic functions of word-forms are also a source of ambiguity. In this paper I shall discuss the points one and two by using Swahili as a test language. 1 Morphological analysis The morphological analysis of Swahili is carried out by SWATWOL, which is based on the two- level formalism (Koskenniemi 1983). The ap- plication of this formalism to Swahili has been under process since 1987, and it has now, af- ter having been tested with a corpus of one mil- lion words, reached a mature phase with a recall of 99.8% in average running text, and precision of close to 100%. The performance of SWAT- WOL corresponds to what is reported of ENGT- WOL, the morphological parser of English (Vouti- lainen et al 1992; Tapanainen and J/irvinen 1994), and SWETWOL, the morphological analyzer of Swedish (Karlsson 1992). SWATWOL uses a two-level rule system for describing morphophonological variation, as well as a lexicon with 288 sub-lexicons. Unlike in languages with right-branching word formation, where word roots can be grouped together into a root lexicon, here word roots have been divided into several sub-lexicons. Because SWATWOL has been described in de- tail elsewhere (Hurskainen 1992), only a sketchy description of its parts is given here. 1.1 SWATWOL rules Two-level rules have been written …

Content truncated to 500 words