Sensitivity of Self-tuning Histograms: Query Order Affecting Accuracy and Robustness

  • Andranik Khachatryan
  • Emmanuel Müller
  • Christian Stier
  • Klemens Böhm
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7338)

Abstract

In scientific databases, the amount and the complexity of data calls for data summarization techniques. Such summaries are used to assist fast approximate query answering or query optimization. Histograms are a prominent class of model-free data summaries and are widely used in database systems.

So-called self-tuning histograms look at query-execution results to refine themselves. An assumption with such histograms is that they can learn the dataset from scratch. We show that this is not the case and highlight a major challenge that stems from this. Traditional self-tuning is overly sensitive to the order of queries, and reaches only local optima with high estimation errors. We show that a self-tuning method can be improved significantly if it starts with a carefully chosen initial configuration. We propose initialization by subspace clusters in projections of the data. This improves both accuracy and robustness of self-tuning histograms.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Baltrunas, L., Mazeika, A., Böhlen, M.H.: Multi-dimensional histograms with tight bounds for the error. In: IDEAS, pp. 105–112 (2006)Google Scholar
  2. 2.
    Beyer, K., Goldstein, J., Ramakrishnan, R., Shaft, U.: When Is ”Nearest Neighbor” Meaningful? In: Beeri, C., Bruneman, P. (eds.) ICDT 1999. LNCS, vol. 1540, pp. 217–235. Springer, Heidelberg (1998)CrossRefGoogle Scholar
  3. 3.
    Bruno, N., Chaudhuri, S., Gravano, L.: Stholes: a multidimensional workload-aware histogram. SIGMOD Rec. 30, 211–222 (2001)CrossRefGoogle Scholar
  4. 4.
    Deshpande, A., Garofalakis, M., Rastogi, R.: Independence is good: dependency-based histogram synopses for high-dimensional data. SIGMOD Rec. 30, 199–210 (2001)CrossRefGoogle Scholar
  5. 5.
    Gunopulos, D., Kollios, G., Tsotras, V.J., Domeniconi, C.: Approximating multi-dimensional aggregate range queries over real attributes. SIGMOD Rec. 29, 463–474 (2000)CrossRefGoogle Scholar
  6. 6.
    Guttman, A.: R-trees: A dynamic index structure for spatial searching. In: Yormark, B. (ed.) SIGMOD 1984, Proceedings of Annual Meeting, Boston, Massachusetts, June 18-21, pp. 47–57. ACM Press (1984)Google Scholar
  7. 7.
    Khachatryan, A., Müller, E., Böhm, K., Kopper, J.: Efficient Selectivity Estimation by Histogram Construction Based on Subspace Clustering. In: Bayard Cushing, J., French, J., Bowers, S. (eds.) SSDBM 2011. LNCS, vol. 6809, pp. 351–368. Springer, Heidelberg (2011)CrossRefGoogle Scholar
  8. 8.
    Luo, J., Zhou, X., Zhang, Y., Shen, H.T., Li, J.: Selectivity estimation by batch-query based histogram and parametric method. In: ADC 2007, pp. 93–102. Australian Computer Society, Inc., Darlinghurst (2007)Google Scholar
  9. 9.
    Müller, E., Günnemann, S., Assent, I., Seidl, T.: Evaluating clustering in subspace projections of high dimensional data. PVLDB 2(1), 1270–1281 (2009)Google Scholar
  10. 10.
    Muthukrishnan, S., Poosala, V., Suel, T.: On Rectangular Partitionings in Two Dimensions: Algorithms, Complexity, and Applications. In: Beeri, C., Bruneman, P. (eds.) ICDT 1999. LNCS, vol. 1540, pp. 236–256. Springer, Heidelberg (1998)CrossRefGoogle Scholar
  11. 11.
    Poosala, V., Ioannidis, Y.E.: Selectivity estimation without the attribute value independence assumption. In: VLDB 1997, pp. 486–495. Morgan Kaufmann Publishers Inc., San Francisco (1997)Google Scholar
  12. 12.
    Roh, Y.J., Kim, J.H., Chung, Y.D., Son, J.H., Kim, M.H.: Hierarchically organized skew-tolerant histograms for geographic data objects. In: SIGMOD 2010, pp. 627–638. ACM, New York (2010)Google Scholar
  13. 13.
    SDSS Collaboration. Sloan Digital Sky Survey (August 27, 2011)Google Scholar
  14. 14.
    Sellis, T.K., Roussopoulos, N., Faloutsos, C.: The r+-tree: A dynamic index for multi-dimensional objects. In: VLDB, pp. 507–518 (1987)Google Scholar
  15. 15.
    Srivastava, U., Haas, P.J., Markl, V., Kutsch, M., Tran, T.M.: Isomer: Consistent histogram construction using query feedback. In: ICDE 2006, p. 39. IEEE Computer Society, Washington, DC (2006)Google Scholar
  16. 16.
    Wang, H., Sevcik, K.C.: A multi-dimensional histogram for selectivity estimation and fast approximate query answering. In: CASCON 2003, pp. 328–342. IBM Press (2003)Google Scholar
  17. 17.
    Wang, H., Sevcik, K.C.: Histograms based on the minimum description length principle. The VLDB Journal 17 (May 2008)Google Scholar
  18. 18.
    Yiu, M.L., Mamoulis, N.: Frequent-pattern based iterative projected clustering. In: ICDM 2003, pp. 689–692 (2003)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Andranik Khachatryan
    • 1
  • Emmanuel Müller
    • 1
  • Christian Stier
    • 1
  • Klemens Böhm
    • 1
  1. 1.Institute for Program Structures and Data Organization (IPD)Karlsruhe Institute of Technology (KIT)Germany

Personalised recommendations