The role of context in automatic speech recognition

Authors: Kokurina N.V., Zhukov D.M.
Published in issue: #4(69)/2022
DOI: 10.18698/2541-8009-2022-4-784
Category: Humanities | Chapter: Social sciences
Keywords: context, speech recognition, speech understanding, micro-context, macro-context, spoken language, lexical context, structural context, syntactic context, cultural context, acoustics
Published: 28.04.2022

The context as a factor affecting the results of automatic speech recognition is considered. The necessity of classifying contexts in automatic natural language processing is described. The structural and non-structural types of contexts are outlined, and the intonational context is highlighted. The lexical and syntactic contexts in which the meaning of a lexical unit depends on the nearest environment are considered. Each of the identified contexts influencing the quality of automatic speech recognition is illustrated with examples and comments. It is emphasized that semantic and situational contexts are of particular importance for automatic speech recognition. The cultural context is highlighted as the most challenging for automatic speech processing.

References

[1] Butenko Yu.I., Shostak I.V. Methodological aspects of automatic speech recognition on the basis of multivariate statistical theory. Neyrokomp’yutery: razrabotka, primenenie [Neurocomputers], 2018, no. 2, pp. 23–33 (in Russ.).

[2] Selivanova E.A. Lingvisticheskaya entsiklopediya Poltava [Poltava linguistic encyclopedia]. Moscow, Dovkilya-K Publ., 2010.

[3] Potapova R.K., Potapov V.V. Rechevaya kommunikatsiya [Verbal communication]. Moscow, Yazyki slavyanskikh kul’tur Publ., 2012 (in Russ.).

[4] Frolov A.V. Sintez i raspoznavanie rechi. Sovremennye resheniya [Synthesis and recognition of speech. Modern solutions]. Moscow, Svyaz’ Publ., 2003 (in Russ.).

[5] Butenko Yu.I. Using trigrams for automatic speech recognition. Vestnik NGU. Seriya: Lingvistika i mezhkul’turnaya kommunikatsiya [NSU Vestnik. Series: Linguistics and Intercultural Communication], 2020, vol. 18, no. 3, pp. 5–15 (in Russ.).

[6] Piotrovskiy R.G. Modelirovanie fonologicheskikh sistem i metody ikh sravneniya [Simulation of phonologic systems and methods for their comparison]. Leningrad, Nauka Publ., 1966 (in Russ.).

[7] Butenko Yu.I., Ermakova Yu.V. [Types pf contexts at oral speech recognition]. XVII Vseros. nauch. konf. Neyrokomp’yutery i ikh primenenie [XVII Russ. Sci. Conf. Neurocomputers and Their Application]. Moscow, MGPPU, 2019, pp. 183–185 (in Russ.).

[8] Kosarev Yu.A., Li I.V., Ronzhin A.L. et al. Review of speech and text understanding methods. Trudy SPIIRAN [SPIIRAS Proceedings], 2002, vol. 2, no. 1, pp. 157–195 (in Russ.).

[9] O ponimanii komp’yuterami teksta [On understanding of text by computers] (in Russ.). URL: https://habr.com/ru/post/126748/ (accessed: 15.08.2019).

[10] Gavrilova T.A., Khoroshevskiy V.F. Bazy znaniy intellektual’nykh system [Knowledge bases of intelligent systems]. Sankt-Petersburg, Piter Publ., 2000 (in Russ.).