Grammar engineering for Swahili

Abstract

Most of the African languages are under resourced languages hence suffer from data sparsity due to lack of sufficient digital corpora making data driven methods not efficient for developing language technology resources. However, the availability of digital devices and ubiquitous computing demands these low-density languages to have language resources for application purposes. Therefore, this paper describes the engineering of Swahili grammar using Grammatical Framework (GF), a rapid grammar writing tool and formalism. A morphology rule based driven approach has been used where morphology is developed first, then followed by the syntactic part. The typical evaluation metrics BLEU and PER metrics were used to evaluate the grammar resulted in encouraging 77.95% and 9.46% respectively. The work is a significant step for the low resourced Swahili language since it provides a morphological analyzer and interlingua machine translation in the GF ecosystem which is useful in the analysis and generation of the language. Finally, the grammar lays a foundation for the development of controlled natural language applications on top of the Swahili grammar and the platform for extracting bilingual corpus for use in data driven methods.