Jointly modeling aspects and opinions with a MaxEnt-LDA hybrid
W. Zhao, J. Jiang, H. Yan, and X. Li. Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, page 56--65. Stroudsburg, PA, USA, Association for Computational Linguistics, (2010)
Abstract
Discovering and summarizing opinions from online reviews is an important and challenging task. A commonly-adopted framework generates structured review summaries with aspects and opinions. Recently topic models have been used to identify meaningful review aspects, but existing topic models do not identify aspect-specific opinion words. In this paper, we propose a MaxEnt-LDA hybrid model to jointly discover both aspects and aspect-specific opinion words. We show that with a relatively small amount of training data, our model can effectively identify aspect and opinion words simultaneously. We also demonstrate the domain adaptability of our model.
%0 Conference Paper
%1 citeulike:9605702
%A Zhao, Wayne X.
%A Jiang, Jing
%A Yan, Hongfei
%A Li, Xiaoming
%B Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing
%C Stroudsburg, PA, USA
%D 2010
%I Association for Computational Linguistics
%K aspects jointly lda maxent opinions
%P 56--65
%T Jointly modeling aspects and opinions with a MaxEnt-LDA hybrid
%U http://portal.acm.org/citation.cfm?id=1870664
%X Discovering and summarizing opinions from online reviews is an important and challenging task. A commonly-adopted framework generates structured review summaries with aspects and opinions. Recently topic models have been used to identify meaningful review aspects, but existing topic models do not identify aspect-specific opinion words. In this paper, we propose a MaxEnt-LDA hybrid model to jointly discover both aspects and aspect-specific opinion words. We show that with a relatively small amount of training data, our model can effectively identify aspect and opinion words simultaneously. We also demonstrate the domain adaptability of our model.
@inproceedings{citeulike:9605702,
abstract = {Discovering and summarizing opinions from online reviews is an important and challenging task. A commonly-adopted framework generates structured review summaries with aspects and opinions. Recently topic models have been used to identify meaningful review aspects, but existing topic models do not identify aspect-specific opinion words. In this paper, we propose a {MaxEnt}-{LDA} hybrid model to jointly discover both aspects and aspect-specific opinion words. We show that with a relatively small amount of training data, our model can effectively identify aspect and opinion words simultaneously. We also demonstrate the domain adaptability of our model.},
added-at = {2012-07-30T13:26:02.000+0200},
address = {Stroudsburg, PA, USA},
author = {Zhao, Wayne X. and Jiang, Jing and Yan, Hongfei and Li, Xiaoming},
biburl = {https://www.bibsonomy.org/bibtex/20067287d27ca7e7d9eab28a04e5d54ab/minghuiqiu},
booktitle = {Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing},
interhash = {16fe66e7f64ee15514cffce2c25263b6},
intrahash = {0067287d27ca7e7d9eab28a04e5d54ab},
keywords = {aspects jointly lda maxent opinions},
location = {Cambridge, Massachusetts},
pages = {56--65},
priority = {2},
publisher = {Association for Computational Linguistics},
series = {EMNLP '10},
timestamp = {2012-07-30T13:26:22.000+0200},
title = {Jointly modeling aspects and opinions with a {MaxEnt}-{LDA} hybrid},
url = {http://portal.acm.org/citation.cfm?id=1870664},
year = 2010
}