Filters








6 Hits in 1.3 sec

Exploiting Explicit and Implicit Feedback for Personalized Ranking

Gai Li, Qiang Chen
2016 Mathematical Problems in Engineering  
and implicit feedback simultaneously and optimized the well-known evaluation metric Expected Reciprocal Rank (ERR).  ...  ++ showed a linear correlation with the number of rating.  ...  The NDCG of all the users is the mean score of each user. ERR is a generalized version of Reciprocal Rank (RR) designed to be used with multiple relevance level data (e.g., ratings).  ... 
doi:10.1155/2016/2535329 fatcat:o7l3meksmbd4tbsaejgyzz2zoa

Top-N-Rank: A Scalable List-wise Ranking Method for Recommender Systems [article]

Junjie Liang, Jinlong Hu, Shoubin Dong, Vasant Honavar
2018 arXiv   pre-print
of implicit feedback with differing levels of reliability or trustworthiness.  ...  We propose Top-N-Rank, a novel family of list-wise Learning-to-Rank models for reliably recommending the N top-ranked items.  ...  Instead of directly optimizing MRR, CLiMF learns the latent factors by maximizing the smoothed lower bound of MRR; (iii) xCLiMF: An extension of CLiMF that optimizes the expected reciprocal rank (ERR),  ... 
arXiv:1812.04109v2 fatcat:nfwnneydkzgazbjdkpkyskwua4

Using graded implicit feedback for bayesian personalized ranking

Lukas Lerche, Dietmar Jannach
2014 Proceedings of the 8th ACM Conference on Recommender systems - RecSys '14  
In the literature, a number of algorithm proposals have been made that rely solely on such implicit feedback, among them Bayesian Personalized Ranking (BPR).  ...  and there can exist several actions for an item over time.  ...  Another learning-to-rang technique is xCLiMF [11] which uses a graded relevance scale and optimizes a rank criterion, in that case the Expected Reciprocal Rank.  ... 
doi:10.1145/2645710.2645759 dblp:conf/recsys/LercheJ14 fatcat:kekhausw7zhtrinrcjir7kb6n4

Collaborative Ranking with a Push at the Top

Konstantina Christakopoulou, Arindam Banerjee
2015 Proceedings of the 24th International Conference on World Wide Web - WWW '15  
From such a perspective, collaborative ranking based formulations with suitable ranking loss functions are natural.  ...  In this paper, building on recent advances from the learning to rank literature, we introduce a novel family of collaborative ranking algorithms which focus on accuracy at the top of the list for each  ...  We also acknowledge technical support from the University of Minnesota Supercomputing Institute.  ... 
doi:10.1145/2736277.2741678 dblp:conf/www/Christakopoulou15 fatcat:ajjqg2h2dzfixhvamg267tsvdm

Using implicit feedback for recommender systems: characteristics, applications, and challenges

Lukas Lerche, Technische Universität Dortmund, Technische Universität Dortmund
2017
Recommender systems are software tools to tackle the problem of information overload by helping users to find items that are most relevant for them within an often unmanageable set of choices.  ...  Using implicit feedback leads to new challenges and open questions regarding, for example, the huge amount of signals to process, the ambiguity of the feedback, and the inevitable noise in the data.  ...  The author of this thesis wrote parts of the text and his specific contributions are the adaptable sampling strategy for BPR and most of the experimentation, analysis, and result interpretation.  ... 
doi:10.17877/de290r-17802 fatcat:oxf6jo4ic5grta7jd2a5tbj2kq

Collaborative ranking-based recommender systems

Jun Hu
2018
In order to generate accurate recommended lists, we look into the technique of combining learning-to-rank with conventional collaborative filtering methods for solving the recommendation task and comprehensively  ...  We view the task of recommendation as providing a personalized ranked list of items for each user and thus formulate it as a ranking problem.  ...  CLiMF [101] and xCLiMF [108] optimize (expected) mean-reciprocal rank (MRR), which has the tendency to obtain at least a few interesting items at the top of ranked list.  ... 
doi:10.7282/t3-27m3-e107 fatcat:a7knah3h4jc6hhbfftzszeh6ka