Abstract
In the context of supervised learning, the training data for large-scale hierarchical classification consist of (i) a set of input-output pairs, and (ii) a hierarchy structure defining parent-child relation among class labels. It is often the case that the hierarchy structure given a-priori is not optimal for achieving high classification accuracy. This is especially true for web-taxonomies such as Yahoo! directory which consist of tens of thousand of classes. Furthermore, an important goal of hierarchy design is to render better navigability and browsing. In this work, we propose a maximum-margin framework for automatically adapting the given hierarchy by using the set of input-output pairs to yield a new hierarchy. The proposed method is not only theoretically justified but also provides a more principled approach for hierarchy flattening techniques proposed earlier, which are ad-hoc and empirical in nature. The empirical results on publicly available large-scale datasets demonstrate that classification with new hierarchy leads to better or comparable generalization performance than the hierarchy flattening techniques.
Access provided by Autonomous University of Puebla. Download to read the full chapter text
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Bennett, P.N., Nguyen, N.: Refined experts: improving classification in large taxonomies. In: Proc. 32nd Int’l ACM SIGIR, pp. 11–18. ACM (2009)
Cai, L., Hofmann, T.: Hierarchical document categorization with support vector machines. In: CIKM, pp. 78–87. ACM (2004)
Dekel, O.: Distribution-calibrated hierarchical classification. In: Advances in Neural Information Processing Systems, pp. 450–458 (2009)
Dekel, O., Keshet, J., Singer, Y.: Large margin hierarchical classification. In: Proceedings of the 21st International Conference on Machine Learning, ICML 2004, pp. 27–34 (2004)
Gao, T., Koller, D.: Discriminative learning of relaxed hierarchy for large-scale visual recognition. In: IEEE International Conference on Computer Vision (ICCV), pp. 2072–2079 (2011)
Liu, T.-Y., Yang, Y., Wan, H., Zeng, H.-J., Chen, Z., Ma, W.-Y.: Support vector machines classification with a very large-scale taxonomy. SIGKDD Explor. Newsl., 36–43 (2005)
Malik, H.: Improving hierarchical svms by hierarchy flattening and lazy classification. In: 1st Pascal Workshop on Large Scale Hierarchical Classification (2009)
Partalas, I., Babbar, R., Gaussier, E., Amblard, C.: Adaptive classifier selection in large-scale hierarchical classification. In: Huang, T., Zeng, Z., Li, C., Leung, C.S. (eds.) ICONIP 2012, Part III. LNCS, vol. 7665, pp. 612–619. Springer, Heidelberg (2012)
Wang, X., Lu, B.-L.: Flatten hierarchies for large-scale hierarchical text categorization. In: Fifth IEEE International Conference on Digital Information Management, pp. 139–144 (2010)
Xue, G.-R., Xing, D., Yang, Q., Yu, Y.: Deep classification in large-scale text hierarchies. In: Proc. 31st Int’l ACM SIGIR, pp. 619–626. ACM
Yang, Y., Liu, X.: A re-examination of text categorization methods. In: Proceedings of the 22nd International ACM SIGIR, pp. 42–49. ACM (1999)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Babbar, R., Partalas, I., Gaussier, E., Amini, MR. (2013). Maximum-Margin Framework for Training Data Synchronization in Large-Scale Hierarchical Classification. In: Lee, M., Hirose, A., Hou, ZG., Kil, R.M. (eds) Neural Information Processing. ICONIP 2013. Lecture Notes in Computer Science, vol 8226. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-42054-2_42
Download citation
DOI: https://doi.org/10.1007/978-3-642-42054-2_42
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-42053-5
Online ISBN: 978-3-642-42054-2
eBook Packages: Computer ScienceComputer Science (R0)