The Pure Cold-Start Problem: A deep study about how to conquer first-time users in recommendations domains

Abstract The success of Web-based applications depends on their ability to convert first-time users into recurring ones. This problem is known as Pure Cold-Start and it refers to the capability of Recommender Systems (RSs) providing useful recommendations to users without historical data. Traditionally, the systems assume that items biased by popularity, recency and positive ratings suit the interests of most first-time users. However, our studies confirm a contra-intuitive hypothesis, showing users consumption preferences biased to non-popular items. For this reason, we introduce two new RSs to mitigate this problem based on user coverage maximization: Max-Coverage and Category-Exploration. Offline experiments show that our recommendations complement the traditional ones. Thus, we found an opportunity for improvements on state-of-the-art RSs. We hypothesize that combining distinct non-personalized RSs can be better to conquer the most first-time users than traditional ones. An online study conducted with 236 real users in movie domain reinforced this hypothesis. Hence, this study provides a clear message: we should compose product pages that mix complementary non-personalized RSs.

[1]  Mi Zhang,et al.  Enhancing diversity in Top-N recommendation , 2009, RecSys '09.

[2]  Sean M. McNee,et al.  Improving recommendation lists through topic diversification , 2005, WWW '05.

[3]  Guy Shani,et al.  Evaluating Recommendation Systems , 2011, Recommender Systems Handbook.

[4]  Fernando Ortega,et al.  A probabilistic model for recommending to new cold-start non-registered users , 2017, Inf. Sci..

[5]  Harald Steck,et al.  Interactive Recommender Systems: Tutorial , 2015, RecSys.

[6]  Vasek Chvátal,et al.  A Greedy Heuristic for the Set-Covering Problem , 1979, Math. Oper. Res..

[7]  Katrien Verbert,et al.  Interactive recommender systems: A survey of the state of the art and future research challenges and opportunities , 2016, Expert Syst. Appl..

[8]  Nava Tintarev,et al.  Joint Workshop on Interfaces and Human Decision Making for Recommender Systems (#IntRS) , 2015, RecSys.

[9]  Saul Vargas,et al.  Rank and relevance in novelty and diversity metrics for recommender systems , 2011, RecSys '11.

[10]  Gediminas Adomavicius,et al.  Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions , 2005, IEEE Transactions on Knowledge and Data Engineering.

[11]  Saul Vargas,et al.  Coverage, redundancy and size-awareness in genre diversity for recommender systems , 2014, RecSys '14.

[12]  Neil J. Hurley,et al.  Novelty and Diversity in Top-N Recommendation -- Analysis and Evaluation , 2011, TOIT.

[13]  Bengt J. Nilsson,et al.  Using maximum coverage to optimize recommendation systems in e-commerce , 2013, RecSys.

[14]  W. Bruce Croft,et al.  Diversity by proportionality: an election-based approach to search result diversification , 2012, SIGIR '12.

[15]  Eduardo R. Hruschka,et al.  Simultaneous co-clustering and learning to address the cold start problem in recommender systems , 2015, Knowl. Based Syst..

[16]  Adam Prügel-Bennett,et al.  Leveraging clustering approaches to solve the gray-sheep users problem in recommender systems , 2014, Expert Syst. Appl..

[17]  Yi-Cheng Zhang,et al.  Solving the apparent diversity-accuracy dilemma of recommender systems , 2008, Proceedings of the National Academy of Sciences.

[18]  Rafael Valencia-García,et al.  Solving the cold-start problem in recommender systems with social tags , 2010, Expert Syst. Appl..

[19]  Angshul Majumdar,et al.  Cold-start, warm-start and everything in between: An autoencoder based approach to recommendation , 2017, 2017 International Joint Conference on Neural Networks (IJCNN).

[20]  Djoerd Hiemstra,et al.  Query recommendation for children , 2012, CIKM '12.

[21]  Saul Vargas,et al.  Intent-oriented diversity in recommender systems , 2011, SIGIR.

[22]  B. Schwartz.  The Paradox of Choice: Why More Is Less , 2004 .

[23]  J. Bobadilla,et al.  Recommender systems survey , 2013, Knowl. Based Syst..

[24]  Hakim Hacid,et al.  Social networks and information retrieval, how are they converging? A survey, a taxonomy and an analysis of social information retrieval approaches and platforms , 2016, Inf. Syst..

[25]  Hayder Radha,et al.  Cold-Start Item and User Recommendation with Decoupled Completion and Transduction , 2015, RecSys.

[26]  Chirayu Wongchokprasitti,et al.  User Model in a Box: Cross-System User Model Transfer for Resolving Cold Start Problems , 2015, UMAP.

[27]  William C. McDowell,et al.  An examination of retail website design and conversion rate , 2016 .

[28]  Akram Salah,et al.  Exploiting User Demographic Attributes for Solving Cold-Start Problem in Recommender System , 2013 .

[29]  Ahmad Nurzid Rosli,et al.  Alleviating the cold-start problem by incorporating movies facebook pages , 2014, Cluster Computing.

[30]  Stathes Hadjiefthymiades,et al.  Facing the cold start problem in recommender systems , 2014, Expert Syst. Appl..

[31]  Yves Grandvalet,et al.  A Coverage-Based Approach to Recommendation Diversity On Similarity Graph , 2016, RecSys.