论文信息 - Dynamic hyperparameter optimization for bayesian topical trend analysis

Dynamic hyperparameter optimization for bayesian topical trend analysis

This paper presents a new Bayesian topical trend analysis. We regard the parameters of topic Dirichlet priors in latent Dirichlet allocation as a function of document timestamps and optimize the parameters by a gradient-based algorithm. Since our method gives similar hyperparameters to the documents having similar timestamps, topic assignment in collapsed Gibbs sampling is affected by timestamp similarities. We compute TFIDF-based document similarities by using a result of collapsed Gibbs sampling and evaluate our proposal by link detection task of Topic Detection and Tracking.

[1] John D. Lafferty,et al. Dynamic topic models , 2006, ICML.

[2] Changhu Wang,et al. Learning to reduce the semantic gap in web image retrieval and annotation , 2008, SIGIR '08.

[3] Chong Wang,et al. Continuous Time Dynamic Topic Models , 2008, UAI.

[4] Sushil Krishna Bajracharya,et al. Mining Internet-Scale Software Repositories , 2007, NIPS.

[5] Jorge Nocedal,et al. On the limited memory BFGS method for large scale optimization , 1989, Math. Program..

[6] Chirag Shah,et al. Representing documents with named entities for story link detection (SLD) , 2006, CIKM '06.

[7] Michael I. Jordan,et al. Latent Dirichlet Allocation , 2001, J. Mach. Learn. Res..

[8] James Allan,et al. UMass at TDT 2002 , 2002 .

[9] Andrew McCallum,et al. Topics over time: a non-Markov continuous-time model of topical trends , 2006, KDD '06.

[10] Ramesh Nallapati,et al. Multiscale topic tomography , 2007, KDD '07.

[11] W. Bruce Croft,et al. LDA-based document models for ad-hoc retrieval , 2006, SIGIR.

[12] Bernt Schiele,et al. Discovery of activity patterns using topic models , 2008 .

[13] Mark Steyvers,et al. Finding scientific topics , 2004, Proceedings of the National Academy of Sciences of the United States of America.

[14] Edward Y. Chang,et al. Combinational collaborative filtering for personalized community recommendation , 2008, KDD.

[15] J. Nocedal. Updating Quasi-Newton Matrices With Limited Storage , 1980 .