“Turklang” Linguistic Knowledge Graph for Creation of Tools for Teaching Turkic Languages

“Turklang” Linguistic Knowledge Graph for Creation of Tools for Teaching Turkic Languages

연구 분야: Databases

학회: Automatic Documentation and Mathematical Linguistics

This article presents elements of the “Turklang” linguistic knowledge graph developed at the Institute of Applied Semiotics of the Academy of Sciences of Tatarstan and used as a basis for creating a number of linguistic resources and tools: the “Turkic Morpheme” portal, the “Tugan Tel” electronic corpus of the Tatar language, linguistic processors. Creating an educational environment requires subject-oriented knowledge graphs, for which methods of general and open graphs are not suitable. This paper describes linguistic knowledge graphs, which reflect, on the one hand, potential capabilities of Turkic languages, and on the other hand, examples of actual use in texts. Peculiarity of these knowledge graphs is that they contain linguistic units of different linguistic levels and concepts corresponding to meanings of these linguistic units, which are built into a single model of the linguistic knowledge graph. Structure of this knowledge graph allows to formulate the content of a training course, build an individual educational trajectory, as well as create tests and tools of automated answer grading as part of knowledge control when teaching Turkic languages. This makes it possible to subsequently develop, based on these graphs, training programs taking into account the structural and functional features of the Turkic languages and also contributes to the implementation of individual goals of students.