Language Models Plus Specialized Documentation: Building Regular Expressions that Match Chinese Grammar Patterns

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Previous research had suggested using authentic teaching materials, corpora and regular expressions in Chinese as a Foreign Language (CFL) courses. Where regular expressions are concerned, however, it can be argued that language teachers are not expected to be experts on language programming or on authoring digital applications when it comes to build them. Given the importance of digital text in the study of additional languages nowadays, this case study discussed the use of language models in combination with specialized documentation to learn about the construction of regular expressions. It is hoped this discussion will be useful for both CFL teachers and learners who are interested in learning to use these expressions. The target language is the shì…de (shì…de (是…的)) grammar patterns. The existence of different shì…de (shì…de (是…的)) grammar patterns made it possible to illustrate the benefits of using language models to build regular expressions that match Chinese grammar patterns. Indeed, in enhancing the accuracy of the technical aspects of regular expressions, language models can also depict nuances in Chinese language text. Nevertheless, these benefits come with challenges, which together with the limitations of this study are tackled in this paper.