International E-Journal of Advances in Social Sciences
Yazarlar: Ksenia BALYSHEVA, Elena KARTASHOVA, Konstantin KONTİEV, Aleksey MİKHEEV
Konular:-
DOI:10.18769/ijasos.336974
Anahtar Kelimeler:Automatic identifying,Morphological properties,Ontology,Ontology-based application,Text analyser,LexInfo,OntoLex
Özet: This article presents the first stage of an ongoing effort of creating the application Ontology-Based Text Analyser (OTA) aimed at automatically identifying semantics and grammatical properties of widely used Russian words in connected texts. At present this application identifies only morphological properties of Russian words. In this application all morphological properties of content words and grammatical function words are revealed on the basis of a query to the Ontology of Russian Grammatical Forms (OntoRuGrammaForm) that we set up earlier. In OntoRuGrammaForm we used LexInfo which represents morphological properties of words in the ontological format as a scheme for data organising. To set up OntoRuGrammaForm the existing LexInfo ontology was extended with missing and refined grammatical categories. In OntoRuGrammaForm the linkage of semantics with morphological properties is implemented with OntoLex which makes it possible to link grammatical word forms with lemmas and lemmas with concepts in knowledge area ontologies. The automatic process of word morphology identification is illustrated with a connected text of the informative type taken from the open news online-portal. In this news text the system of morphological properties of words is identified with OntoRuGrammaForm. This application also displays lemmas and transcription of separate words in a connected text. The created application (OTA) can be used as an innovative methodical tool in teaching Russian to foreign students to develop skills of identifying morphological characteristics of words in texts. At present this application is available on the Web in open access and can be used for analysing morphological properties of widely used Russian words in a connected text.