Utterance classification using linguistic and non-linguistic information for network-based speech-to-speech translation systems

Komei Sugiura, Ryong Lee, Hideki Kashioka, Koji Zettsu, Yutaka Kidawara

Research output: Contribution to journalConference article

Abstract

Network-based mobile services, such as speech-to-speech translation and voice search, enable the construction of large-scale log database including speech. We have developed a smartphone application called VoiceTra for speech-to-speech translation and have collected 10, 000, 000 utterances so far. This huge corpus is unique in size and spatio-temporal information, it contains information on anonymized user locations. This spatio-temporal corpus can be used for improving the accuracy of its speech recognition and machine translation, and it will open the door for the study of the location dependency of vocabulary and new applications for location-based services. This paper first analyzes the corpus and then presents a novel method for classifying utterances using linguistic and non-linguistic information. L2-regularized Logistic Regression is used for utterance classification. Our experiments performed on the VoiceTra log corpus revealed that our proposed method outperformed baseline methods in terms of F measure.

Original languageEnglish
Article number6569092
Pages (from-to)212-216
Number of pages5
JournalProceedings - IEEE International Conference on Mobile Data Management
Volume2
DOIs
Publication statusPublished - 2013 Sep 11
Externally publishedYes
Event14th International Conference on Mobile Data Management, MDM 2013 - Milan, Italy
Duration: 2013 Jun 32013 Jun 6

Keywords

  • GIS
  • smartphone
  • speech-to-speech translation

ASJC Scopus subject areas

  • Engineering(all)

Fingerprint Dive into the research topics of 'Utterance classification using linguistic and non-linguistic information for network-based speech-to-speech translation systems'. Together they form a unique fingerprint.

  • Cite this