Problems of Semantic Relation Extraction from Tatar Text Corpus “Tugan Tel”

This study is an intermediate result of work on the task of automatic taxonomy enrichment by semantic relations of hyponymy-hypernymy of Turkic languages multilingual thesaurus, implemented within the framework of the linguistic portal “Turkic Morpheme”. Data extraction for the thesaurus is carried out from the electronic corpus of Tatar texts “Tugan Tel”. The article analyzes possible methods for extracting semantic relations, taking into account the difficulties associated with their implementation for Tatar language processing. To this end, the most effective solutions proposed within the RUSSE'2020 competition for hypernymy extraction, as well as already implemented approaches for the Turkic languages, are considered. The list of resources required to implement these methods for the Tatar language is analyzed. Based on the analysis, a brief description of the available electronic resources for the Tatar language is given. As a result, several actual problems with the task of extracting semantic relations for the Tatar language are highlighted.