22 November 2021

KILS-2021: Meeting of the section “Computer and corpus linguistics”

The meeting was held on November 15 from 11.00 to 13.30. It was attended by 20 speakers and listeners in the Microsoft Teams team. 13 reports were presented. Geography of the authors of the reports: 2 countries are represented – Russia and Israel, 8 cities (St. Petersburg – 5 people, Moscow – 4, Kazan – 4, Izhevsk – 2, Krasnoyarsk – 2, Saransk – 1, Yekaterinburg – 1, Chelyabinsk – 1) … Among the authors of reports there are 2 doctors of sciences, 8 candidates of sciences.

The reports of the section will touch upon a wide range of topical topics. The following reports can be highlighted. In the report of Kolmogorova A.V. and Vdovina L.A. “Emotional Analysis of VKontakte Social Network Texts: Experience of a Linguistic Project Based on Machine Learning” presents an approach to describing fuzzy (non-discrete) emotions using machine learning methods, including with the help of modern deep learning neural networks. Report by E.V. Beloglazova “What the Headlines Talk About: A Corpus Study of the ROSSICA Bibliography” is devoted to the analysis of the corpus of foreign publications about Russia, which makes it possible to understand how Russia is perceived abroad. Report by A.N. Laposhina, M.Yu. Lebedeva. “It depends on how to count: the influence of the corpus type on the data on the frequency of a word in the context of determining the complexity of educational texts for primary school” is devoted to the problem of creating frequency dictionaries based on the material of school textbooks. The creation of such dictionaries is important because they can be used, in particular, to determine the level of difficulty of textbooks. In the report of Evgrafova K.V. and Sokolova N.S. “Acoustic analysis of the quality of voice and speech of foreign language teachers in higher education during the period of online learning in a pandemic” shows an increase in the load on the teacher’s vocal apparatus when using distance technologies. In the report of A.V. Zeldin. an innovative parametric model of phonemes is proposed, which allows for quantitative studies of the degree of phonetic proximity of words. In general, it should be noted the high level of all the reports presented, the wide use of various corpora and tools for mathematical analysis of linguistic data.