While Grammatical Framework (GF) resource grammars are primarily intended as linguistic software libraries, they can also be employed to enable syntax parsing via the GF runtime system. The isiZulu resource grammar (ZRG) in conjunction with the GF runtime system constitutes the first openly available parser for isiZulu. A key feature of the implementation of the ZRG is to model aspects of morphology as equivalent to syntax. In the absence of a baseline, we evaluate the parser at the token level against a morphological analyser, before characterising its phrase level analysis in order to establish a baseline for parsing of isiZulu text.
Reference:
Marais, L. & Pretorius, L. 2023. Parsing IsiZulu text using grammatical framework. Lecture Notes in Networks and Systems, 741. http://hdl.handle.net/10204/13373
Marais, L., & Pretorius, L. (2023). Parsing IsiZulu text using grammatical framework. Lecture Notes in Networks and Systems, 741, http://hdl.handle.net/10204/13373
Marais, Laurette, and L Pretorius "Parsing IsiZulu text using grammatical framework." Lecture Notes in Networks and Systems, 741 (2023) http://hdl.handle.net/10204/13373
Marais L, Pretorius L. Parsing IsiZulu text using grammatical framework. Lecture Notes in Networks and Systems, 741. 2023; http://hdl.handle.net/10204/13373.