Learning rules and categorization networks for language standardization

Abstract

In this research, the authors use machine learning techniques to provide solutions for descriptive linguists in the domain of language standardization. With regard to the personal name construction in Afrikaans, the authors perform function learning from word pairs using the Default and Refine algorithm. The authors demonstrate how the extracted rules can be used to identify irregularities in previously standardized constructions and to predict new forms of unseen words. In addition, the authors defined a generic, automated process that allows them to extract constructional schemas and present these visually as categorization networks, similar to what is often being used in Cognitive Grammar. The authors conclude that computational modeling of constructions can contribute to new descriptive linguistic insights, and to practical language solutions.