Abstract
We introduce perturbation kernels, a new class of similarity measure for information retrieval that casts word similarity in terms of multi-task learning. Perturbation kernels model uncertainty in the user’s query by choosing a small number of variations in the relative weights of the query terms to build a more complete picture of the query context, which is then used to compute a form of expected distance between words. Our approach has a principled mathematical foundation, a simple analytical form, and makes few assumptions about the underlying retrieval model, making it easy to apply in a broad family of existing query expansion and model estimation algorithms.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Ando, R.K., Dredze, M., Zhang, T.: TREC 2005 genomics track experiments at IBM Watson. In: Proceedings of TREC 2005, NIST Special Publication (2006)
Ando, R.K., Zhang, T.: A framework for learning predictive structures from multiple tasks and unlabeled data. J. Mach. Learning Research 6, 1817–1853 (2005)
Baxter, J.: The canonical distortion measure for vector quantization and function approximation. In: ICML 1997, pp. 39–47 (1997)
Collins-Thompson, K.: Estimating robust query models using convex optimization. In: Advances in Neural Information Processing Systems (NIPS), vol. 21, pp. 329–336. MIT Press, Cambridge (2008)
Collins-Thompson, K.: Robust Model Estimation Methods for Information Retrieval, PhD thesis. Carnegie Mellon University (2008)
Dillon, J., Mao, Y., Lebanon, G., Zhang, J.: Statistical translation, heat kernels, and expected distances. In: UAI 2007, pp. 93–100 (2007)
Jaakkola, T., Haussler, D.: Exploiting generative models in discriminative classifiers. In: Advances in Neural Information Processing Systems (NIPS), vol. 11, pp. 487–493. MIT Press, Cambridge (1999)
Jebara, T., Kondor, R., Howard, A.: Probability product kernels. J. Machine Learning Research 5, 819–844 (2004)
Lafferty, J.D., Lebanon, G.: Information diffusion kernels. In: Advances in Neural Information Processing Systems (NIPS), vol. 15, pp. 375–382. MIT Press, Cambridge (2002)
Lavrenko, V.: A Generative Theory of Relevance. PhD thesis, Univ. of Massachusetts, Amherst (2004)
Minka, T.: Distance measures as prior probabilities. Technical report (2000)
Sahami, M., Heilman, T.: A web-based kernel function for measuring the similarity of short text snippets. In: Proc. of WWW 2006, pp. 377–386 (2006)
Tsuda, K., Kawanabe, M.: The leave-one-out kernel. In: Dorronsoro, J.R. (ed.) ICANN 2002. LNCS, vol. 2415, pp. 727–732. Springer, Heidelberg (2002)
Tsuda, K., Kin, T., Asai, K.: Marginalized kernels for biological sequences. Bioinformatics 18(1), 268–275 (2002)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Collins-Thompson, K. (2009). Robust Word Similarity Estimation Using Perturbation Kernels. In: Azzopardi, L., et al. Advances in Information Retrieval Theory. ICTIR 2009. Lecture Notes in Computer Science, vol 5766. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-04417-5_25
Download citation
DOI: https://doi.org/10.1007/978-3-642-04417-5_25
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-04416-8
Online ISBN: 978-3-642-04417-5
eBook Packages: Computer ScienceComputer Science (R0)