Journal of the Korean Data and Information Science Society

The Korean Data and Information Science Society (한국데이터정보과학회)
권호 표지
DOI QR코드

DOI QR Code

Revisiting the Bradley-Terry model and its application to information retrieval

  • Received : 2013.07.01
  • Accepted : 2013.09.16
  • Published : 2013.09.30
Download PDF
⟨ Previous Next ⟩

Abstract

The Bradley-Terry model is widely used for analysis of pairwise preference data. We explain that the popularity of Bradley-Terry model is gained due to not only easy computation but also some nice asymptotic properties when the model is misspecified. For information retrieval required to analyze big ranking data, we propose to use a pseudo likelihood based on the Bradley-Terry model even when the true model is different from the Bradley-Terry model. We justify using the Bradley-Terry model by proving that the estimated ranking based on the proposed pseudo likelihood is consistent when the true model belongs to the class of Thurstone models, which is much bigger than the Bradley-Terry model.

Keywords

References

  1. Bartlett, P. L., Jordan, M. I. and Mcauliffe, J. D. (2006). Convextity, classification, and risk bounds. Journal of the American Statistical Association, 101, 138-156. https://doi.org/10.1198/016214505000000907
  2. Berk, N. K. (1973). A central limit theorem for m-dependent random variables with unbouned m. Annals of Probability, 1, 352-354. https://doi.org/10.1214/aop/1176996992
  3. Bradley, R. A. (1972). A biometrics invited paper. science, statistics, and paired comparisons. Biometrcs, 39, 213-239.
  4. Bradley, R. A. and Terry, M. E. (1952). Rank analysis of incomplete block designs: I. The method of paired comparisons. Biometrika, 39, 324-345.
  5. Burges, C., Shaked, T., Renshaw, E., Lazier, A., Deeds, M., Hamilton, N. and Hullender, G. (2005). Learning to rank using gradient descent. Proceedings of the 22nd International Conference on Machine Learning, 89-96.
  6. Critchlow E. E. and Flinger M. A. (1991). Paired comparison, triple comparison, and ranking experiments as generalized linear models, and their implementation on GLIM. Psychometirka, 56, 517-533. https://doi.org/10.1007/BF02294488
  7. Critchlow E. E., Flinger M. A. and Verducci J. S. (1991). Probability models on rankings. Journal of Mathematical Psychology, 35, 294-318. https://doi.org/10.1016/0022-2496(91)90050-4
  8. Fligner, M. A. and Verducci, J. S. (1986). Distance based ranking models. Journal of the Royal Statistical Society B, 48, 359-369.
  9. Freund, Y., Iyer, R., Schapire, R. E. and Singer, Y. (2003). An efficient boosting algorithm for combining preferences. The Journal of Machine Learning Research, 4, 933-963.
  10. Hong C., Jung M. and Lee J. (2010). Prediction model analysis of 2010 South Africa World Cup. Journal of the Korean Data & Information Science Society, 21, 1137-1146.
  11. Joachims, T. (2003). Optimizing search engines using clickthrough data. Proceedings of the 8th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 133-142.
  12. Li K-C. and Duan N. (1989). Regression analysis under link violation. Annals of Statistics, 17, 1009-1052. https://doi.org/10.1214/aos/1176347254
  13. Luce, R. D. (1959). Individual choice behavior: A theoretical analysis, John Wiley, New York, 213-239.
  14. Mallows, C. L. (1957). Non-null ranking models. I. Biometrika, 44, 114-130. https://doi.org/10.1093/biomet/44.1-2.114
  15. Plackett, R. L. (1975). The analysis of permutations. Applied Statistics, 24, 193{202. https://doi.org/10.2307/2346567
  16. Pollard, D. (1991). Asymptotics for least absolute deviation regression estimators. Econometric Theory, 7, 186-199. https://doi.org/10.1017/S0266466600004394
  17. Pregibon, D. (1980). Goodness of link tests for generalized linear models. Applied Statistics, 29, 15-24. https://doi.org/10.2307/2346405
  18. Stern, H. (1990). Models for distributions on permutations. Journal of the American Statistical Association, 85, 558-564. https://doi.org/10.1080/01621459.1990.10476235
  19. Thurstone, L. L. (1927). A law of comparative judgment. Psychological Review, 44, 273-286.
  20. Xia, F., Liu, T-Y., Wang, J., Zhang, W. and Li, H. (2008). Listwise approach to learning to rank: Theory and algorithm, Proceedings of the 25th International Conference on Machine Learning, 1192-1199.
  21. Yan, T., Yang, Y. and Xu, J. (2010). Sparse paired comparisons in the Bradley-Terry model. Statistica Sinica, 22, 1305-1318.
  22. Zhang, T. (2004). Statistical behavior and consistency of classification methods based on convex risk mini- mization. Annals of Statistics, 32, 56-134.

Cited by

  1. The accuracy of farmer-generated data in an agricultural citizen science methodology vol.37, pp.4, 2017, https://doi.org/10.1007/s13593-017-0441-y