Learning to select for a predefined ranking

Aleksei Ustimenko, Aleksandr Vorobev, Gleb Gusev, Pavel Serdyukov
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:6477-6486, 2019.

Abstract

In this paper, we formulate a novel problem of learning to select a set of items maximizing the quality of their ordered list, where the order is predefined by some explicit rule. Unlike the classic information retrieval problem, in our setting, the predefined order of items in the list may not correspond to their quality in general. For example, this is a dominant scenario in personalized news and social media feeds, where items are ordered by publication time in a user interface. We propose new theoretically grounded algorithms based on direct optimization of the resulting list quality. Our offline and online experiments with a large-scale product search engine demonstrate the overwhelming advantage of our methods over the baselines in terms of all key quality metrics.

Cite this Paper

BibTeX
@InProceedings{pmlr-v97-vorobev19a, title = {Learning to select for a predefined ranking}, author = {Ustimenko, Aleksei and Vorobev, Aleksandr and Gusev, Gleb and Serdyukov, Pavel}, booktitle = {Proceedings of the 36th International Conference on Machine Learning}, pages = {6477--6486}, year = {2019}, editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan}, volume = {97}, series = {Proceedings of Machine Learning Research}, month = {09--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v97/vorobev19a/vorobev19a.pdf}, url = {https://proceedings.mlr.press/v97/vorobev19a.html}, abstract = {In this paper, we formulate a novel problem of learning to select a set of items maximizing the quality of their ordered list, where the order is predefined by some explicit rule. Unlike the classic information retrieval problem, in our setting, the predefined order of items in the list may not correspond to their quality in general. For example, this is a dominant scenario in personalized news and social media feeds, where items are ordered by publication time in a user interface. We propose new theoretically grounded algorithms based on direct optimization of the resulting list quality. Our offline and online experiments with a large-scale product search engine demonstrate the overwhelming advantage of our methods over the baselines in terms of all key quality metrics.} }
Endnote
%0 Conference Paper %T Learning to select for a predefined ranking %A Aleksei Ustimenko %A Aleksandr Vorobev %A Gleb Gusev %A Pavel Serdyukov %B Proceedings of the 36th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Ruslan Salakhutdinov %F pmlr-v97-vorobev19a %I PMLR %P 6477--6486 %U https://proceedings.mlr.press/v97/vorobev19a.html %V 97 %X In this paper, we formulate a novel problem of learning to select a set of items maximizing the quality of their ordered list, where the order is predefined by some explicit rule. Unlike the classic information retrieval problem, in our setting, the predefined order of items in the list may not correspond to their quality in general. For example, this is a dominant scenario in personalized news and social media feeds, where items are ordered by publication time in a user interface. We propose new theoretically grounded algorithms based on direct optimization of the resulting list quality. Our offline and online experiments with a large-scale product search engine demonstrate the overwhelming advantage of our methods over the baselines in terms of all key quality metrics.
APA
Ustimenko, A., Vorobev, A., Gusev, G. & Serdyukov, P.. (2019). Learning to select for a predefined ranking. Proceedings of the 36th International Conference on Machine Learning, in Proceedings of Machine Learning Research 97:6477-6486 Available from https://proceedings.mlr.press/v97/vorobev19a.html.

Related Material