DSpace university logo mark
Advanced Search
Japanese | English 

NAOSITE : Nagasaki University's Academic Output SITE > Faculty of Engineering > Articles in academic journal >

Steering time-dependent estimation of posteriors with hyperparameter indexing in Bayesian topic models

File Description SizeFormat
LNCS6634_435.pdf168.56 kBAdobe PDFView/Open

Title: Steering time-dependent estimation of posteriors with hyperparameter indexing in Bayesian topic models
Authors: Masada, Tomonari / Takasu, Atsuhiro / Shibata, Yuichiro / Oguri, Kiyoshi
Issue Date: 2011
Publisher: Springer Verlag
Citation: Lecture Notes in Computer Science, 6634 LNAI(PART 1), pp.435-447; 2011
Abstract: This paper provides a new approach to topical trend analysis. Our aim is to improve the generalization power of latent Dirichlet allocation (LDA) by using document timestamps. Many previous works model topical trends by making latent topic distributions time-dependent. We propose a straightforward approach by preparing a different word multinomial distribution for each time point. Since this approach increases the number of parameters, overfitting becomes a critical issue. Our contribution to this issue is two-fold. First, we propose an effective way of defining Dirichlet priors over the word multinomials. Second, we propose a special scheduling of variational Bayesian (VB) inference. Comprehensive experiments with six datasets prove that our approach can improve LDA and also Topics over Time, a well-known variant of LDA, in terms of test data perplexity in the framework of VB inference.
Keywords: Bayesian methods / parallelization / topic models / trend analysis / variational inference
URI: http://hdl.handle.net/10069/25516
ISSN: 03029743
Rights: © 2011 Springer-Verlag. / The original publication is available at www.springerlink.com
Type: Journal Article
Text Version: author
Appears in Collections:Articles in academic journal

Citable URI : http://hdl.handle.net/10069/25516

All items in NAOSITE are protected by copyright, with all rights reserved.


Valid XHTML 1.0! Copyright © 2006-2015 Nagasaki University Library - Feedback Powerd by DSpace