1932

Abstract

Preference data occur when assessors express comparative opinions about a set of items, by rating, ranking, pair comparing, liking, or clicking. The purpose of preference learning is to () infer on the shared consensus preference of a group of users, sometimes called rank aggregation, or () estimate for each user her individual ranking of the items, when the user indicates only incomplete preferences; the latter is an important part of recommender systems. We provide an overview of probabilistic approaches to preference learning, including the Mallows, Plackett–Luce, and Bradley–Terry models and collaborative filtering, and some of their variations. We illustrate, compare, and discuss the use of these methods by means of an experiment in which assessors rank potatoes, and with a simulation. The purpose of this article is not to recommend the use of one best method but to present a palette of different possibilities for different questions and different types of data.

Loading

Article metrics loading...

/content/journals/10.1146/annurev-statistics-031017-100213
2019-03-07
2024-04-17
Loading full text...

Full text loading...

/deliver/fulltext/statistics/6/1/annurev-statistics-031017-100213.html?itemId=/content/journals/10.1146/annurev-statistics-031017-100213&mimeType=html&fmt=ahah

Literature Cited

  1. Adomavicius G, Tuzhilin A. 2015. Context-aware recommender systems. Recommender Systems Handbook F Ricci, L Rokach, B Shapira, PB Kantor191–226 New York: Springer
    [Google Scholar]
  2. Agarwal DK, Chen BC 2016. Statistical Methods for Recommender Systems Cambridge, UK: Cambridge Univ. Press
  3. Aggarwal CC 2016. Recommender Systems New York: Springer
  4. Agresti A 1996. Categorical Data Analysis New York: Wiley
  5. Alvo M, Yu PLH 2014. Statistical Methods for Ranking Data New York: Springer
  6. Asfaw D, Vitelli V, Sørensen Ø, Arjas E, Frigessi A. 2017. Time-varying rankings with the Bayesian Mallows model. Stat 6:14–30
    [Google Scholar]
  7. Bradley RA, Terry ME. 1952. Rank analysis of incomplete block designs: I. The method of paired comparisons. Biometrika 39:324–45
    [Google Scholar]
  8. Caron F, Doucet A. 2012. Efficient Bayesian inference for generalized Bradley–Terry models. J. Comput. Graph. Stat. 21:174–96
    [Google Scholar]
  9. Caron F, Teh YW, Murphy TB. 2014. Bayesian nonparametric Plackett-Luce models for the analysis of preferences for college degree programmes. Ann. Appl. Stat. 8:1145–81
    [Google Scholar]
  10. Cattelan M. 2012. Models for paired comparison data: a review with emphasis on dependent data. Stat. Sci. 27:412–33
    [Google Scholar]
  11. Crispino M, Vitelli V, Barrett N, Arjas E, Frigessi A. 2017. A Bayesian Mallows approach to non-transitive pair comparison data: How human are sounds?. arXiv:1705.08805 [stat.AP]
    [Google Scholar]
  12. Davidson RR. 1970. On extending the Bradley–Terry model to accommodate ties in paired comparison experiments. J. Am. Stat. Assoc. 65:317–28
    [Google Scholar]
  13. Desarkar MS, Sarkar S, Mitra P. 2016. Preference relations based unsupervised rank aggregation for metasearch. Expert Syst. Appl. 49:86–98
    [Google Scholar]
  14. Diaconis P 1988. Group Representations in Probability and Statistics Hayward, CA: Inst. Math. Stat.
  15. Diaconis P, Graham RL. 1977. Spearman's footrule as a measure of disarray. J. R. Stat. Soc. B 39:262–68
    [Google Scholar]
  16. Fang H, Wang Y, Jin Q, Ma J. 2017. RankwithTA: A robust and accurate peer grading mechanism for MOOCs. 2017 IEEE 6th International Conference on Teaching, Assessment, and Learning for Engineering (TALE)497–502 Washington, DC: IEEE
    [Google Scholar]
  17. Ford LR. 1957. Solution of a ranking problem from binary comparisons. Am. Math. Mon. 64:28–33
    [Google Scholar]
  18. Fürnkranz J, Hüllermeier E 2010. Preference Learning: An Introduction New York: Springer
  19. Glickman ME. 1999. Parameter estimation in large dynamic paired comparison experiments. Appl. Stat. 48:377–94
    [Google Scholar]
  20. Green P, Hastie T. 1999. Reversible jump MCMC. Genetics 155:1391–1403
    [Google Scholar]
  21. Gunawardana A, Shani G. 2015. Evaluating recommender systems. Recommender Systems Handbook F Ricci, L Rokach, B Shapira, PB Kantor265–308 New York: Springer
    [Google Scholar]
  22. Hu Y, Koren Y, Volinsky C. 2008. Collaborative filtering for implicit feedback datasets. Proceedings of the 2008 Eighth IEEE International Conference on Data Mining (ICDM ’08) Washington, DC: IEEE
    [Google Scholar]
  23. Hunter DR. 2004. MM algorithms for generalized Bradley-Terry models. Ann. Stat. 32:384–406
    [Google Scholar]
  24. Irurozki E, Calvo B, Lozano A. 2014. Sampling and learning the Mallows and generalized Mallows models under the Hamming distance Tech. Rep., Dep. Comput. Sci. Artif. Intell., Univ. Basque Ctry., Leioa, Spain
  25. Irurozki E, Calvo B, Lozano JA. 2018. Sampling and learning Mallows and generalized Mallows models under the Cayley distance. Methodol. Comput. Appl. Probab. 20:1–35
    [Google Scholar]
  26. Jacques J, Biernacki C. 2014. Model-based clustering for multivariate partial ranking data. J. Stat. Plann. Inference 149:201–17
    [Google Scholar]
  27. Karatzoglou A, Hidasi B. 2017. Deep learning for recommender systems. Proceedings of the Eleventh ACM Conference on Recommender Systems396–97 New York: ACM
    [Google Scholar]
  28. Kemeny JG, Snell JL 1962. Mathematical Models in the Social Sciences Waltham, MA: Blaisdell
  29. Koren Y, Bell R, Volinsky C. 2009. Matrix factorization techniques for recommender systems. Computer 42:30–37
    [Google Scholar]
  30. Lansdowne ZF, Woodward BS. 1996. Applying the Borda ranking method. Air Force J. Logist. 20:27–29
    [Google Scholar]
  31. Li L, Chu W, Langford J, Schapire RE. 2010. A contextual-bandit approach to personalized news article recommendation. Proceedings of the 19th International Conference on World Wide Web661–70 New York: ACM
    [Google Scholar]
  32. Lu T, Boutilier C. 2014. Effective sampling and learning for Mallows models with pairwise-preference data. J. Mach. Learn. Res. 15:3783–829
    [Google Scholar]
  33. Luce RD 1959. Individual Choice Behavior: A Theoretical Analysis New York: Wiley
  34. Mallows CL. 1957. Non-null ranking models. I. Biometrika 44:114–30
    [Google Scholar]
  35. Marden JI. 1995. Analyzing and Modeling Rank Data Boca Raton, FL: CRC/Chapman & Hall
  36. Meil M, Chen H. 2010. Dirichlet Process Mixtures of Generalized Mallows Models. Proceedings of the Twenty-Sixth Conference Annual Conference on Uncertainty in Artificial Intelligence (UAI-10) Corvallis, OR: AUAI
    [Google Scholar]
  37. Mollica C, Tardella L 2014. Epitope profiling via mixture modeling for ranked data. Stat. Med. 33:213738–58
    [Google Scholar]
  38. Mollica C, Tardella L 2017. Bayesian Plackett–Luce mixture models for partially ranked data. Psychometrika 82:2442–58
    [Google Scholar]
  39. Mukherjee S. 2016. Estimation in exponential families on permutations. Ann. Stat. 44:853–75
    [Google Scholar]
  40. Plackett RL. 1975. The analysis of permutations. J. R. Stat. Soc. C 24:193–202
    [Google Scholar]
  41. Raman K, Joachims T. 2015. Bayesian ordinal peer grading. Proceedings of the Second (2015) ACM Conference on Learning @ Scale149–56 New York: ACM
    [Google Scholar]
  42. Rao P, Kupper LL. 1967. Ties in paired-comparison experiments: a generalization of the Bradley–Terry model. J. Am. Stat. Assoc. 62:194–204
    [Google Scholar]
  43. Ricci F, Rokach L, Shapira B. 2015. Recommender systems: introduction and challenges. Recommender Systems Handbook F Ricci, L Rokach, B Shapira, PB Kantor1–34 New York: Springer
    [Google Scholar]
  44. Salakhutdinov R, Mnih A. 2008. Bayesian probabilistic matrix factorization using Markov chain Monte Carlo. Proceedings of the 25th International Conference on Machine Learning (ICML ’08) New York: ACM
    [Google Scholar]
  45. Tanner M, Wong W. 1987. The calculation of posterior distributions by data augmentation (with discussion). J. Am. Stat. Assoc. 82:528–50
    [Google Scholar]
  46. Thurstone LL. 1927. A law of comparative judgment. Psychol. Rev. 34:273
    [Google Scholar]
  47. Turner H, van Etten J, Firth D, Kosmidis I. 2018. Introduction to PlackettLuce. https://cran.rstudio.com/web/packages/PlackettLuce/vignettes/Overview.html
    [Google Scholar]
  48. Vitelli V, Sørensen Ø, Crispino M, Frigessi A, Arjas E. 2018. Probabilistic preference learning with the Mallows rank model. J. Mach. Learn. Res. 18:1–49
    [Google Scholar]
  49. Wu R, Xu J, Srikant R, Massoulié L, Lelarge M, Hajek B. 2015. Clustering and inference from pairwise comparisons. Proceedings of the 2015 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems (SIGMETRICS ’15) New York: ACM
    [Google Scholar]
  50. Yan T. 2016. Ranking in the generalized Bradley–Terry models when the strong connection condition fails. Commun. Stat. Theory Methods 45:340–53
    [Google Scholar]
  51. Zermelo E. 1929. Die Berechnung der Turnier-Ergebnisse als ein Maximumproblem der Wahrscheinlichkeitsrechnung. Math. Z. 29:436–60
    [Google Scholar]
/content/journals/10.1146/annurev-statistics-031017-100213
Loading
/content/journals/10.1146/annurev-statistics-031017-100213
Loading

Data & Media loading...

  • Article Type: Review Article
This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error