Topic Number Estimation by Consensus Soft Clustering with NMF

  • Takeru Yokoi
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6485)


We propose here a novel method to estimate the number of topics in a document set using consensus clustering based on Non-negative Matrix Factorization (NMF). It is useful to automatically estimate the number of topics from a document set since various approaches to extract topics can determine their number through heuristics. Consensus clustering makes it possible to obtain a consensus of multiple results of clustering so that robust clustering is achieved and the number of clusters is regarded as the optimized number. In this paper, we have proposed a novel consensus soft clustering algorithm based on NMF and estimated an optimized number of topics by searching through a robust classification of documents for the topics obtained.


Consensus Clustering Estimation of the number of topics Soft Clustering Topic extraction 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Larsen, B., Aone, C.: Fast and Effective Text Mining using Linear-time Document Clustering. In: 5th International Conference on Knowledge Discovery and Data Mining (SIGKDD), pp. 16–22 (1999)Google Scholar
  2. 2.
    Pelleg, D., Moore, A.: X-means: Extending K-means with Efficient Estimation of the Number of Clusters. In: 17th International Conference on Machine Learning, pp. 727–734 (2000)Google Scholar
  3. 3.
    Windham, M., Culter, A.: Information Ratios for Validating Mixture Analysis. Journal of the American Statistical Association 87, 1182–1192 (1992)CrossRefGoogle Scholar
  4. 4.
    The, Y.W., Jordan, M.I., Beal, M.J., Blei, D.M.: Hierarchical Dirichlet Process. Technical Report 653, Department of Statistics, University of California at Berkeley (2004)Google Scholar
  5. 5.
    Monti, S., Tamayo, P., Mesirov, J., Golub, T.: Consensus Clustering: A Resampling-Based Method for Class Discovery and Visualization of Gene Expression Microarray Data. Journal of Machine Learning 52, 91–118 (2003)CrossRefzbMATHGoogle Scholar
  6. 6.
    Li, T., Ding, C.: Weighted Consensus Clustering. In: Jonker, W., Petković, M. (eds.) SDM 2008. LNCS, vol. 5159, pp. 798–809. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  7. 7.
    Brunet, J.P., Tamayo, P., Golub, T.R., Mesirov, J.P.: Metagenes and Molecular Pattern Discovery using Matrix Factorization. PNAS 101(12), 4164–4169 (2004)CrossRefGoogle Scholar
  8. 8.
    Rui, X., Wunsch II, D.C.: Clustering, pp. 267–268. J. Wiley & Sons Inc., NJ (2009)Google Scholar
  9. 9.
    Berry, M.W., Browne, M., Langville, A.N.: Algorithms and Applications for Approximate Nonnegative Matrix Factorization, V. In: Pauca, V.P., Plemmons, R.J. (eds.) Computational Statistics & Data Analysis, vol. 52(1), pp. 155–173 (2008)Google Scholar
  10. 10.
    Salton, G., McGill, M.J.: Introduction to Modern Information Retrieval. McGraw-Hill Book Company, New York (1983)zbMATHGoogle Scholar
  11. 11.
    Lee, D.D., Seung, H.S.: Algorithms for Non-negative Matrix Factorization. Advanced Neural Information Processing Systems 13, 556–562 (2001)Google Scholar
  12. 12.
    Punera, K., Ghosh, J.: Consensus-Based Ensembles of Soft Clustering. In: International Conference on Machine Learning: Models, Technologies & Applications (MLMTA 2007), pp. 3–9 (2007)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Takeru Yokoi
    • 1
  1. 1.Tokyo Metropolitan College of Industrial TechnologyJapan

Personalised recommendations