Time-aware conversion prediction

Abstract

The importance of product recommendation has been well recognized as a central task in business intelligence for e-commerce websites. Interestingly, what has been less aware of is the fact that different products take different time periods for conversion. The “conversion” here refers to actually a more general set of pre-defined actions, including for example purchases or registrations in recommendation and advertising systems. The mismatch between the product’s actual conversion period and the application’s target conversion period has been the subtle culprit compromising many existing recommendation algorithms.

The challenging question: what products should be recommended for a given time period to maximize conversion—is what has motivated us in this paper to propose a rank-based time-aware conversion prediction model (rTCP), which considers both recommendation relevance and conversion time. We adopt lifetime models in survival analysis to model the conversion time and personalize the temporal prediction by incorporating context information such as user preference. A novel mixture lifetime model is proposed to further accommodate the complexity of conversion intervals. Experimental results on two real-world data sets illustrate the high goodness of fit of our proposed model rTCP and demonstrate its effectiveness in time-aware conversion rate prediction for advertising and product recommendation.

Keywords

conversion time / survival analysis / product recommendation / advertising

References

[1] YuJ J, ZhuT Y. Combining long-term and short-term user interest for personalized hashtag recommendation. Frontiers of Computer Science, 2015, 9(4): 608–622

[2] DingY, LiX. Time weight collaborative filtering. In: Proceedings of the 14th ACM International Conference on Information and Knowledge Management. 2005, 485–492

[3] CamposP G, DiezF, CantadorI. Time-aware recommender systems: a comprehensive survey and analysis of existing evaluation protocols. User Modeling and User-Adapted Interaction, 2014, 24(1): 67–119

[4] YuanQ, CongG, MaZ Y, Sun A, ThalmannN M . Time-aware pointof- interest recommendation. In: Proceedings of the 36th International ACMSIGIR Conference on Research and Development in Information Retrieval. 2013, 363–372

[5] BannurS, AlonsoO. Analyzing temporal characteristics of check-in data. In: Proceedings of the Companion Publication of the 23rd International Conference on World Wide Web Companion. 2014, 827–832

[6] PazzaniM, Billsus D. Learning and revising user profiles: the identification of interesting web sites. Machine Learning, 1997, 27(3): 313–331

[7] ZhuT S, Greiner R, HaublG . Learning a model of a web user’s interests. In: Proceedings of the 2003 International Conference on User Modeling. 2003, 65–75

[8] PazzaniM J. A framework for collaborative, content-based and demographic filtering. Artificial Intelligence Review, 1999, 13(5): 393–408

[9] GuanY, CaiS M, ShangM S. Recommendation algorithm based onitem quality and user rating preferences. Frontiers of Computer Science, 2014, 8(2): 289–297

[10] GoldbergK, RoederT, GuptaD, Perkins C. Eigentaste: a constant time collaborative filtering algorithm. Information Retrieval, 2001, 4(2): 133–151

[11] BrownP F, PietraV J D, MercerR L, Pietra S A D, LaiJ C . An estimate of an upper bound for the entropy of English. Computational Linguistics, 1992, 18(1): 31–40

[12] BillsusD, Pazzani M J. Learning Collaborative Information Filters. In: Proceedings of the 15th International Conference on Machine Learning. 1998, 46–54

[13] MaH, LiuC, KingI, Lyu M R. Probabilistic factor models for Web site recommendation. In: Proceedings of the 34th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2011, 265–274

[14] BreeseJ S, Heckerman D, KadieC . Empirical analysis of predictive algorithms for collaborative filtering. In: Proceedings of the 14th Conference on Uncertainty in Artificial Intelligence. 1998, 43–52

[15] BleiD M, NgA Y, JordanM I. Latent Dirichlet allocation. Journal of Machine Learning Research, 2003, 3: 993–1022

[16] RendleS, Freudenthaler C, Schmidt-ThiemeL . Factorizing personalized markov chains for next-basket recommendation. In: Proceedings of the 19th International Conference on World Wide Web. 2010, 811–820

[17] ChengC, YangH Q, LyuM R, King I. Where you like to go next: successive point-of-interest recommendation. In: Proceedings of the 23rd International Joint Conference on Artificial Intelligence. 2013, 2605–2611

[18] RosalesR, ChengH, ManavogluE. Post-click conversion modeling and analysis for non-guaranteed delivery display advertising. In: Proceedings of the 5th ACM International Conference on Web Search and Data Mining. 2012, 293–302

[19] ChapelleO, Manavoglu E, RosalesR . Simple and scalable response prediction for display advertising. ACM Transactions on Intelligent Systems and Technology, 2015, 5(4)

[20] LeeK C, OrtenB, DasdanA, Li W J. Estimating conversion rate in display advertising from past erformance data. In: Proceedings of the 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2012, 768–776

[21] LathiaN, HailesS, CapraL, Amatriain X. Temporal diversity in recommender systems. In: Proceedings of the 33rd International ACM SIGIR Conference on Research and Development in Information Retrieval. 2010, 210–217

[22] XiangL, YuanQ, ZhaoS W, Chen L, ZhangX T , YangQ, SunJ M. Temporal recommendation on graphs via long- and short-term preference fusion. In: Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and DataMining. 2010, 723–732

[23] KorenY. Collaborative filtering with temporal dynamics. Communications of the ACM, 2010, 53(4): 89–97

[24] ZhaoG, LeeM L, HsuW, Chen W. Increasing temporal diversity with purchase intervals. In: Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Information Retrieval. 2012, 165–174

[25] GuW R, DongS B, ZengZ Z. Increasing recommended effectiveness with markov chains and purchase intervals. Neural Computing and Applications, 2014, 25(5): 1153–1162

[26] ZhangY F, ZhangM, ZhangY, Lai G K, LiuY Q , ZhangH H, MaS P. Daily-aware personalized recommendation based on feature-level time series analysis. In: Proceedings of the 24th International Conference on World Wide Web. 2015, 1373–1383

[27] ZhaoX W, GuoY W, HeY L, Jiang H, WuY X , LiX M. We know what you want to buy: a demographic-based system for product recommendation on microblogs. In: Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2014, 1935–1944

[28] NelsonW B. Applied Life Data Analysis. New York:John Wiley & Sons, 2005

[29] ChapelleO. Modeling delayed feedback in display advertising. In: Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Aining. 2014, 1097–1105

[30] WangJ, ZhangY. Opportunity model for e-commerce recommendation: right product; right time. In: Proceedings of the 36th International ACMSIGIR Conference on Research and Development in Information Retrieval. 2013, 303–312

[31] WangJ, ZhangY, PosseC, Bhasin A. Is it time for a career switch? In: Proceedings of the 22nd International Conference on World Wide Web. 2013, 1377–1388

[32] RichardsS J. A handbook of parametric survival models for actuarial use. Scandinavian Actuarial Journal, 2012, 2012(4): 233–257

[33] MurphyK P. Machine Learning: A Probabilistic Perspective. Cambridge, Massathusetts: The MIT press, 2012

[34] BarbieriN, MancoG. An analysis of probabilistic methods for top-n recommendation in collaborative filtering. In: Proceedings of Joint European Conference on Machine Learning and Knowledge Discovery in Databases. 2011, 172–187

RIGHTS & PERMISSIONS

2016 Higher Education Press and Springer-Verlag Berlin Heidelberg