Multi-view semi-supervised learning for dialog act segmentation of speech

Basit Öğe Kaydı
dc.authorid0000-0002-4597-0954
dc.authorid0000-0001-5246-2117
dc.contributor.authorGüz, Ümiten_US
dc.contributor.authorCuendet, Sebastienen_US
dc.contributor.authorHakkani Tür, Dileken_US
dc.contributor.authorTür, Gökhanen_US
dc.date.accessioned2015-01-15T23:01:36Z
dc.date.available2015-01-15T23:01:36Z
dc.date.issued2010-02
dc.departmentIşık Üniversitesi, Mühendislik Fakültesi, Elektrik-Elektronik Mühendisliği Bölümüen_US
dc.departmentIşık University, Faculty of Engineering, Department of Electrical-Electronics Engineeringen_US
dc.descriptionThis material is based upon work supported in part by the Defense Advanced Research Projects Agency (DARPA) CALO (FA8750-07-D-0185, Delivery Order 0004), in part by the Scientific and Technological Research Council of Turkey (TUBITAK) funding at SRI, in part by a J. William Fulbright Post-Doctoral Research Fellowship, Isik University Research Fund ( Projects: 05B304, 09A301), DARPA GALE (HR0011-06-C-0023) and in part by the Swiss National Science Foundation through the research network, IM2 fundings at ICSI. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the funding agencies. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Ruhi Sarikayaen_US
dc.description.abstractSentence segmentation of speech aims at determining sentence boundaries in a stream of words as output by the speech recognizer. Typically, statistical methods are used for sentence segmentation. However, they require significant amounts of labeled data, preparation of which is time-consuming, labor-intensive, and expensive. This work investigates the application of multi-view semi-supervised learning algorithms on the sentence boundary classification problem by using lexical and prosodic information. The aim is to find an effective semi-supervised machine learning strategy when only small sets of sentence boundary-labeled data are available. We especially focus on two semi-supervised learning approaches, namely, self-training and co-training. We also compare different example selection strategies for co-training, namely, agreement and disagreement. Furthermore, we propose another method, called self-combined, which is a combination of self-training and co-training. The experimental results obtained on the ICSI Meeting (MRDA) Corpus show that both multi-view methods outperform self-training, and the best results are obtained using co-training alone. This study shows that sentence segmentation is very appropriate for multi-view learning since the data sets can be represented by two disjoint and redundantly sufficient feature sets, namely, using lexical and prosodic information. Performance of the lexical and prosodic models is improved by 26% and 11% relative, respectively, when only a small set of manually labeled examples is used. When both information sources are combined, the semi-supervised learning methods improve the baseline F-Measure of 69.8% to 74.2%.en_US
dc.description.sponsorshipWilliam Fulbright Post-Doctoral Research Fellowshipen_US
dc.description.sponsorshipIsik Universityen_US
dc.description.sponsorshipUnited States Department of Defense
dc.description.sponsorshipSwiss National Science Foundation (SNSF)en_US
dc.description.versionPublisher's Versionen_US
dc.identifier.citationGüz, Ü., Cuendet, S., Hakkani Tür, D. & Tür, G. (2010). Multi-view semi-supervised learning for dialog act segmentation of speech. IEEE Transactions on Audio, Speech, and Language Processing, 16(2), 320-329. doi:10.1109/TASL.2009.2028371en_US
dc.identifier.doi10.1109/TASL.2009.2028371
dc.identifier.endpage329
dc.identifier.issn1558-7916
dc.identifier.issn1558-7924
dc.identifier.issue2
dc.identifier.scopus2-s2.0-85008565008
dc.identifier.scopusqualityQ1
dc.identifier.startpage320
dc.identifier.urihttps://hdl.handle.net/11729/378
dc.identifier.urihttp://dx.doi.org/10.1109/TASL.2009.2028371
dc.identifier.volume18
dc.identifier.wosWOS:000271967900005
dc.identifier.wosqualityQ2
dc.indekslendigikaynakWeb of Scienceen_US
dc.indekslendigikaynakScopusen_US
dc.indekslendigikaynakScience Citation Index Expanded (SCI-EXPANDED)en_US
dc.institutionauthorGüz, Ümiten_US
dc.institutionauthorid0000-0002-4597-0954
dc.language.isoenen_US
dc.peerreviewedYesen_US
dc.publicationstatusPublisheden_US
dc.publisherIEEE-INST Electrical Electronics Engineers Incen_US
dc.relation.ispartofIEEE Transactions on Audio, Speech, and Language Processingen_US
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanıen_US
dc.rightsinfo:eu-repo/semantics/closedAccessen_US
dc.subjectBoostingen_US
dc.subjectCo-trainingen_US
dc.subjectProsodyen_US
dc.subjectSelf-trainingen_US
dc.subjectSemi-supervised learningen_US
dc.subjectSentence segmentationen_US
dc.titleMulti-view semi-supervised learning for dialog act segmentation of speechen_US
dc.typeArticleen_US
dspace.entity.typePublication

Dosyalar

Orijinal paket
Listeleniyor 1 - 1 / 1
Küçük Resim Yok
İsim:
378.pdf
Boyut:
988.84 KB
Biçim:
Adobe Portable Document Format
Açıklama:
Publisher's Version
İndir

Koleksiyon

Makale Koleksiyonu | Elektrik-Elektronik Mühendisliği Bölümü
Scopus İndeksli Yayınlar Koleksiyonu
WoS İndeksli Yayınlar Koleksiyonu