Advertisement

Knowledge and Information Systems

, Volume 31, Issue 3, pp 475–503 | Cite as

Sequential latent Dirichlet allocation

  • Lan Du
  • Wray Buntine
  • Huidong Jin
  • Changyou Chen
Regular Paper

Abstract

Understanding how topics within a document evolve over the structure of the document is an interesting and potentially important problem in exploratory and predictive text analytics. In this article, we address this problem by presenting a novel variant of latent Dirichlet allocation (LDA): Sequential LDA (SeqLDA). This variant directly considers the underlying sequential structure, i.e. a document consists of multiple segments (e.g. chapters, paragraphs), each of which is correlated to its antecedent and subsequent segments. Such progressive sequential dependency is captured by using the hierarchical two-parameter Poisson–Dirichlet process (HPDP). We develop an efficient collapsed Gibbs sampling algorithm to sample from the posterior of the SeqLDA based on the HPDP. Our experimental results on patent documents show that by considering the sequential structure within a document, our SeqLDA model has a higher fidelity over LDA in terms of perplexity (a standard measure of dictionary-based compressibility). The SeqLDA model also yields a nicer sequential topic structure than LDA, as we show in experiments on several books such as Melville’s ‘Moby Dick’.

Keywords

Latent Dirichlet allocation Poisson–Dirichlet process Collapsed Gibbs sampler Topic model Document structure 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Ahmed A, Xing EP (2010) Timeline: a dynamic hierarchical Dirichlet process model for recovering birth/death and evolution of topics in text stream. In: Proceedings of the twenty-sixth conference annual conference on uncertainty in artificial intelligence’Google Scholar
  2. 2.
    AlSumait L, Barbará D, Domeniconi C (2008) On-line LDA: adaptive topic models for mining text streams with applications to topic detection and tracking. In: Proceedings of the eighth international conference on data mining, pp 3–12Google Scholar
  3. 3.
    Blei D, Lafferty J (2006a) Correlated topic models. In: Advances in neural information processing systems, vol 18, pp 147–154Google Scholar
  4. 4.
    Blei DM, Griffiths TL, Jordan MI (2010) The nested Chinese restaurant process and Bayesian nonparametric inference of topic hierarchies. J ACM 57(2): 1–30MathSciNetCrossRefGoogle Scholar
  5. 5.
    Blei DM, Lafferty JD (2006b) Dynamic topic models. In: Proceedings of the 23rd international conference on Machine learning, pp 113–120Google Scholar
  6. 6.
    Blei DM, Ng AY, Jordan MI (2003) Latent Dirichlet allocation. J Mach Learn Res 3: 993–1022zbMATHGoogle Scholar
  7. 7.
    Blei D, McAuliffe J (2007) Supervised topic models. In: Advances in neural information processing systems, vol 20, pp 121–128Google Scholar
  8. 8.
    Buntine W, Du L, Nurmi P (2010) Bayesian networks on Dirichlet distributed vectors. In: Proceedings of the fifth European workshop on probabilistic graphical models (PGM-2010), pp 33–40Google Scholar
  9. 9.
    Buntine W, Hutter M (2010) A bayesian review of the poisson–dirichlet process, Technical Report arXiv:1007.0296, NICTA and ANU, Australia. http://arxiv.org/abs/1007.0296
  10. 10.
    Buntine W, Jakulin A (2006) Discrete components analysis, In: Subspace, latent structure and feature selection techniques. Springer, BerlinGoogle Scholar
  11. 11.
    Du L, Buntine W, Jin H (2010) A segmented topic model based on the two-parameter Poisson–Dirichlet process. Mach Learn 81: 5–19CrossRefGoogle Scholar
  12. 12.
    Du L, Buntine WL, Jin H (2010b) Sequential latent Dirichlet allocation: discover underlying topic structures within a document. In: Proceedings of the 2010 IEEE international conference on data mining. ICDM ’10, pp 148–157Google Scholar
  13. 13.
    Gilks WR, Wild P (1992) Adaptive rejection sampling for Gibbs sampling. J R Stat Soc Ser C 41(2): 337–348zbMATHGoogle Scholar
  14. 14.
    Griffiths TL, Steyvers M (2004) Finding scientific topics. Proc Natl Acad Sci USA 101(1): 5228–5235CrossRefGoogle Scholar
  15. 15.
    Griffiths TL, Steyvers M, Blei DM, Tenenbaum JB (2005) Integrating topics and syntax. In: Advances in neural information processing systems, vol 17, pp 537–544Google Scholar
  16. 16.
    He Q, Chen B, Pei J, Qiu B, Mitra P, Giles L (2009) Detecting topic evolution in scientific literature: how can citations help?. In: Proceeding of the 18th ACM conference on information and knowledge management, pp 957–966Google Scholar
  17. 17.
    Hofmann T (1999) Probabilistic latent semantic indexing. In: Proceedings of the 22nd annual international ACM SIGIR conference on research and development in information retrieval. ACM, pp 50–57Google Scholar
  18. 18.
    Ishwaran H, James LF (2001) Gibbs sampling methods for stick breaking priors. J Am Stat Assoc 96: 161–173MathSciNetzbMATHCrossRefGoogle Scholar
  19. 19.
    Kandylas V, Upham S, Ungar L (2008) Finding cohesive clusters for analyzing knowledge communities. Knowl Inform Syst 17: 335–354CrossRefGoogle Scholar
  20. 20.
    Li T (2008) Clustering based on matrix approximation: a unifying view. Knowl Inform Syst 17: 1–15zbMATHCrossRefGoogle Scholar
  21. 21.
    Mimno D, McCallum A (2008) Topic models conditioned on arbitrary features with Dirichlet-multinomial regression. In: Proceedings of the twenty-fourth conference annual conference on uncertainty in artificial intelligence, pp 411–418Google Scholar
  22. 22.
    Minka TP (2000) Estimating a Dirichlet distribution. Technical report, MITGoogle Scholar
  23. 23.
    Nallapati RM, Ditmore S, Lafferty JD, Ung K (2007) Multiscale topic tomography. In: Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining, pp 520–529Google Scholar
  24. 24.
    Newman D, Asuncion A, Smyth P, Welling M (2008) Distributed inference for latent Dirichlet allocation. In: Advances in neural information processing systems, vol 20, pp 1081–1088Google Scholar
  25. 25.
    Peng W, Li T (2011) Temporal relation co-clustering on directional social network and author-topic evolution. Knowl Inform Syst 26: 467–486MathSciNetCrossRefGoogle Scholar
  26. 26.
    Pitman J, Yor M (1997) The two-parameter Poisson–Diriclet distribution derived from a stable subordinator. Ann Prob 25(2): 855–900MathSciNetzbMATHCrossRefGoogle Scholar
  27. 27.
    Porteous I., Newman D., Ihler A., Asuncion A., Smyth P, Welling M (2008) Fast collapsed Gibbs sampling for latent Dirichlet allocation. In: Proceeding of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining, pp 569–577Google Scholar
  28. 28.
    Ren L, Dunson DB, Carin L (2008) The dynamic hierarchical dirichlet process. In: Proceedings of the 25th international conference on machine learning, pp 824–831Google Scholar
  29. 29.
    Rosen-Zvi M, Griffiths T, Steyvers M, Smyth P (2004) The author-topic model for authors and documents. In: Proceedings of the 20th conference on Uncertainty in Artificial Intelligence. AUAI Press, Arlington, Virginia, United States, pp 487–494Google Scholar
  30. 30.
    Shafiei MM, Milios EE (2006) Latent Dirichlet co-clustering. In: Proceedings of the sixth international conference on data mining, pp 542–551Google Scholar
  31. 31.
    Shen ZY, Sun J, Shen YD (2008) Collective latent Dirichlet allocation. In: Proceedings of the 2008 eighth IEEE international conference on data mining. IEEE Computer Society, Washington, DC, pp 1019–1024Google Scholar
  32. 32.
    Teh Y (2006a) A Bayesian interpretation of interpolated Kneser-Ney, Technical Report TRA2/06, School of Computing. National University of SingaporeGoogle Scholar
  33. 33.
    Teh YW (2006b) A hierarchical Bayesian language model based on Pitman-Yor processes. In: Proceedings of the 21st international conference on computational linguistics and the 44th annual meeting of the association for computational linguistics, pp 985–992Google Scholar
  34. 34.
    Teh Y, Jordan MI, Beal MJ, Blei DM (2006) Hierarchical Dirichlet processes. J Am Stat Assoc 101Google Scholar
  35. 35.
    Thurau C, Kersting K, Wahabzada M, Bauckhage C (2010) Convex non-negative matrix factorization for massive datasets. Knowl Inform Syst. doi: 10.1007/s10115-010-0352-6
  36. 36.
    Wang C, Blei D, Heckerman D (2008) Continuous time dynamic topic models. In: Proceedings of the 24th annual conference on uncertainty in artificial intelligence, pp 579–586Google Scholar
  37. 37.
    Wang H, Huang M, Zhu X (2008) A generative probabilistic model for multi-label classification. In: Proceedings of the 2008 eighth IEEE international conference on data mining. IEEE Computer Society, Washington, DC, pp 628–637Google Scholar
  38. 38.
    Wang X, McCallum A (2006) Topics over time: a non-Markov continuous-time model of topical trends. In: Proceedings of the 12th ACM SIGKDD international conference on knowledge discovery and data mining, pp 424–433Google Scholar
  39. 39.
    Wei X, Sun J, Wang X (2007) Dynamic mixture models for multiple time series. In: Proceedings of the 20th international joint conference on artifical intelligence. Morgan Kaufmann Publishers Inc., pp 2909–2914Google Scholar
  40. 40.
    Zhang J, Song Y, Zhang C, Liu S (2010) Evolutionary hierarchical dirichlet processes for multiple correlated time-varying corpora. In: Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining, pp 1079–1088Google Scholar

Copyright information

© Springer-Verlag London Limited 2011

Authors and Affiliations

  • Lan Du
    • 1
    • 2
  • Wray Buntine
    • 1
    • 2
  • Huidong Jin
    • 1
    • 3
  • Changyou Chen
    • 1
    • 2
  1. 1.CECS, The Australian National UniversityCanberraAustralia
  2. 2.National ICT AustraliaCanberraAustralia
  3. 3.CSIRO Mathematics, Informatics and StatisticsCanberraAustralia

Personalised recommendations