SWS Academic Research eLibraryEarth & Planetary Sciences

Scholarly record

TRAINING ASPECTS OF AUTOMATIC SPEECH RECOGNITION SYSTEMS DURING CHAT BOT CREATION

Mikhail Belenko

First published: 2019-06-20https://doi.org/10.5593/sgem2019/2.1/s07.089View metrics

Abstract

This article describes the key features of the automatic speech recognition systems training process that can be used in a work of chat bot or voice assistant. Importance of research in this area is grazed. Difference between chat bot, voice helper and voice assistance are explained. Basic requirements for correct chat bot usage are formulated. Some common automatic speech recognition systems are reviewed. Kaldi and CMU Sphinx has been used as an open source systems? example. The whole installation and training process are reviewed, instruction for further researchers are documented. Examined systems have been trained with different parameter sets and training data. As a result, the most important training parameters of automatic speech recognition system for chat bot are designated and described. Criteria for selecting data for proper ASR testing is formed. Peculiar chat bot prototype development has started. Results of this research can help to significantly improve the speed and quality of the speech recognition in chat bots and voice assistants. Next steps to continue work with personal solution are presented.

Publication Impact Profile

PlumX
  • Captures
  • Mendeley - Readers: 2
PLUMX see details

Publication details

Title
TRAINING ASPECTS OF AUTOMATIC SPEECH RECOGNITION SYSTEMS DURING CHAT BOT CREATION
Authors
Mikhail Belenko
Proceedings
SGEM International Multidisciplinary Scientific GeoConference EXPO Proceedings; 19th International Multidisciplinary Scientific GeoConference SGEM2019, Informatics, Geoinformatics and Remote Sensing
Publisher
STEF92 Technology
Year
2019
Pages
681-688
SWS Citekey
Belenko20197681688
ISSN
1314-2704
ISBN
978-619-7408-79-9
Language
en
Publication type
Conference Paper
Keywords
chat botspeech recognitiontrainingKaldiCMU Sphinx
References0
0references registered for this publication

Structured references will appear here after the reference import pass. The count is preserved now so the scholarly record is not incomplete.

Citing literature

Number of times cited according to Crossref: 1

Open DOICrossref metadataOpenCitations
View or Download full articleAccess options
Full paper accessChoose SWS login, librarian support, or instant article download.

SWS access login

Login as SWS Scientific Committee

Authors and approved SWS contributors will read and export their own linked papers after identity matching by SWS profile, email and SGEM GlobalID.

For librarian assistance: [email protected]

Purchase Instant Access

48-hour online accessComing soon
Online-only accessComing soon
Download the full article in PDF formatEUR 35
  • Article can be downloaded after successful payment.
  • Article may be used according to SWS library access terms.
  • Article cannot be redistributed.
Get full paper

Back to publication list