Finding Outlying Items in Sets of Partial Rankings
Partial rankings are totally ordered subsets of a set of items. For example, the sequence in which a user browses through different parts of a website is a partial ranking. We consider the following problem. Given a set D of partial rankings, find items that have strongly different status in different parts of D. To do this, we first compute a clustering of D and then look at items whose average rank in the cluster substantially deviates from its average rank in D. Such items can be seen as those that contribute the most to the differences between the clusters. To test the statistical significance of the found items, we propose a method that is based on a MCMC algorithm for sampling random sets of partial rankings with exactly the same statistics as D. We also demonstrate the method on movie rankings and gene expression data.
KeywordsRandom Data Partial Ranking MCMC Algorithm Swap Operation Pair Order
- 3.Kamishima, T., Akaho, S.: Efficient clustering for orders. In: ICDMW 2006. Proceedings of the Sixth IEEE International Conference on Data Mining - Workshops, pp. 274–278. IEEE Computer Society Press, Los Alamitos (2006)Google Scholar
- 5.Scherf, U., Ross, D.T., Waltham, M., Smith, L.H., Lee, J.K., Tanabe, L., Kohn, K.W., Reinhold, W.C., Myers, T.G., TAndrews, D., Scudiero, D.A., Eisen, M.B., Sausville, E.A., Pommier, Y., Botstein, D., Brown, P.O., Weinstein, J.N.: A gene expression database for the molecular pharmacology of cancer. Nature Genetics 24(3), 236–244 (2000)CrossRefGoogle Scholar
- 6.Ukkonen, A., Mannila, H.: Finding representative sets of bucket orders from partial rankings (submitted for review)Google Scholar