Abstract
This paper presents a useful approach for making topic words distribution more accurate and ranking topic significance according to the Jensen-Shannon divergence from background topic as a post-procedure of LDA method. In this paper, at first we defined the term score parameter to represent topics that will suppress the correlation between different topics and make the word distribution more accurate. Then according to the correlation between different topics, we described a concrete method for determining the proper setting of the number of topics. After that we proposed a method for ranking topic significance in the order of the Jensen-Shannon divergence from background topic. As a confirmation of our proposed methods, we conducted several experiments to processing English Twitter streaming data. The results of these experiments validate that our methods work efficiently as expected.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Blei, D.M., Ng, A.Y., Jordan, M.I.: Latent Dirichlet allocation. J. Mach. Learn. Res. 3, 993–1022 (2003)
AlSumait, L., Barbará, D., Gentle, J., Domeniconi, C.: Topic significance ranking of LDA generative Models. In: Buntine, W., Grobelnik, M., Mladenić, D., Shawe-Taylor, J. (eds.) ECML PKDD 2009, Part I. LNCS, vol. 5781, pp. 67–82. Springer, Heidelberg (2009)
Wang, L., Wei, B., Yuan, J.: Topic discovery based on LDA_col model and topic sinificance re-ranking. J. Comput. 6(8), 1639–1647 (2011)
Hofmann, T. : Probabilistic latent sematic analysis. In: UAI, pp. 289–296 (1999)
Baeza-Yates, R., Ribeiro-Neto, B.: Modern Information Retrieval – the Concepts and Technology behind Search, 2nd edn. Pearson Education Limited, Harlow (2011)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Fujino, I., Hoshino, Y. (2015). Making Topic Words Distribution More Accurate and Ranking Topic Significance According to the Jensen-Shannon Divergence from Background Topic. In: Perner, P. (eds) Advances in Data Mining: Applications and Theoretical Aspects. ICDM 2015. Lecture Notes in Computer Science(), vol 9165. Springer, Cham. https://doi.org/10.1007/978-3-319-20910-4_14
Download citation
DOI: https://doi.org/10.1007/978-3-319-20910-4_14
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-20909-8
Online ISBN: 978-3-319-20910-4
eBook Packages: Computer ScienceComputer Science (R0)