MODEL ADAPTATION FOR DIALOG ACT TAGGING Export

@inproceedings{citeulike:3406876, abstract = {In this paper, we analyze the effect of model adaptation for dialog act tagging. The goal of adaptation is to improve the performance of the tagger using out-of-domain data or models. Dialog act tagging aims to provide a basis for further discourse analysis and understanding in conversational speech. In this study we used the ICSI meeting corpus with high-level meeting recognition dialog act (MRDA) tags, that is, question, statement, backchannel, disruptions, and floor grabbers/holders. We performed controlled adaptation experiments using the Switchboard (SWBD) corpus with SWBD-DAMSL tags as the out-of-domain corpus. Our results indicate that we can achieve significantly better dialog act tagging by automatically selecting a subset of the Switchboard corpus and combining the confidences obtained by both in-domain and out-of-domain models via logistic regression, especially when the in-domain data is limited.}, author = {Tur, G. and Guz, U. and Hakkani-Tur, D.}, booktitle = {Spoken Language Technology Workshop, 2006. IEEE}, citeulike-article-id = {3406876}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/SLT.2006.326825}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4123370}, day = {19}, doi = {10.1109/SLT.2006.326825}, journal = {Spoken Language Technology Workshop, 2006. IEEE}, keywords = {act, dialog}, month = {March}, pages = {94--97}, posted-at = {2009-10-01 18:59:45}, priority = {0}, title = {MODEL ADAPTATION FOR DIALOG ACT TAGGING}, url = {http://dx.doi.org/10.1109/SLT.2006.326825}, year = {2007} }

TY - CONF ID - citeulike:3406876 L3 - citeulike-article-id:3406876 N2 - In this paper, we analyze the effect of model adaptation for dialog act tagging. The goal of adaptation is to improve the performance of the tagger using out-of-domain data or models. Dialog act tagging aims to provide a basis for further discourse analysis and understanding in conversational speech. In this study we used the ICSI meeting corpus with high-level meeting recognition dialog act (MRDA) tags, that is, question, statement, backchannel, disruptions, and floor grabbers/holders. We performed controlled adaptation experiments using the Switchboard (SWBD) corpus with SWBD-DAMSL tags as the out-of-domain corpus. Our results indicate that we can achieve significantly better dialog act tagging by automatically selecting a subset of the Switchboard corpus and combining the confidences obtained by both in-domain and out-of-domain models via logistic regression, especially when the in-domain data is limited. JF - Spoken Language Technology Workshop, 2006. IEEE EP - 97 TI - MODEL ADAPTATION FOR DIALOG ACT TAGGING T2 - Spoken Language Technology Workshop, 2006. IEEE SP - 94 KW - act KW - dialog AU - Tur, G AU - Guz, U AU - Hakkani-Tur, D PY - 2007/03/19/ UR - http://dx.doi.org/10.1109/SLT.2006.326825 ER -