The current paper aims to investigate the perception of Russian ironic laboratory speech and to develop a methodological basis for the further research on perceptually relevant acoustic cues of irony. On the first stage of analysis we created short monologues and dialogues including ironic and homonymous non-ironic sentences. Two Russian native male speakers read this material. The recording was accomplished in a sound booth. Then, randomized fragments of the recordings were suggested to native listeners in an auditory perception experiment. The mean intensity, duration of the stressed vowel and F0 range in the reliably identified target fragments were analyzed, as well as functional intonation models in terms of the system suggested by E. Bryzgunova. The results showed that the listeners are able to detect irony without any contextual or lexical marker of it. The acoustic analysis of the reliably identified stimuli demonstrated the difference in intensity level and stressed vowel duration between ironic and non-ironic utterances. The obtained results allow creating a methodology of data collection and validation for the construction of ironic speech corpus and its analysis. They may also have applications in teaching Russian as a second language, human-machine communication and improving natural language processing systems.

Original languageEnglish
Title of host publicationSpeech and Computer - 22nd International Conference, SPECOM 2020, Proceedings
EditorsAlexey Karpov, Rodmonga Potapova
PublisherSpringer Nature
Pages544-553
Number of pages10
ISBN (Print)9783030602758
DOIs
StatePublished - 2020
Event22nd International Conference on Speech and Computer - St. Petersburg, Russia => Online, St. Petersburg, Russian Federation
Duration: 7 Oct 20209 Oct 2020
http://specom.nw.ru/2020/program/SPECOM-ICR2020-Conference-Program-06102020.pdf

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12335 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference22nd International Conference on Speech and Computer
Abbreviated titleSPECOM 2020
Country/TerritoryRussian Federation
CitySt. Petersburg
Period7/10/209/10/20
Internet address

    Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

    Research areas

  • Acoustic analysis, Auditory perception experiment, Intensity, Intonation models, Irony, Phonetics, Pitch

ID: 70314230