Paper Session 6: Algorithms II

Date: Tuesday, Aug 29, 2017, 11:00-12:30
Location: Room 1
Chair: George Karypis

  • LPExpediting Exploration by Attribute-to-Feature Mapping for Cold-Start Recommendations by Deborah Cohen, Michal Aharon, Yair Koren, Oren Somekh and Raz Nissim

    The item cold-start problem is inherent to collaborative filtering (CF) recommenders where items and users are represented by vectors in a latent space. It emerges since CF recommenders rely solely on historical user interactions to characterize their item inventory. As a result, an effective serving of new and trendy items to users may be delayed until enough user feedback is received, thus, reducing both users’ and content suppliers’ satisfaction.

    To mitigate this problem, many commercial recommenders apply random exploration and devote a small portion of their traffic to explore new items and gather interactions from random users. Alternatively, content or context information is combined into the CF recommender, resulting in a hybrid system. Another hybrid approach is to learn a mapping between the item attribute space and the CF latent feature space, and use it to characterize the new items providing initial estimates for their latent vectors.

    In this paper, we adopt the attribute-to-feature mapping approach to expedite random exploration of new items and present LearnAROMA – an advanced algorithm for learning the mapping, previously proposed in the context of classification. In particular, LearnAROMA learns a Gaussian distribution over the mapping matrix. Numerical evaluation demonstrates that this learning technique achieves more accurate initial estimates than logistic regression methods. We then consider a random exploration setting, in which new items are further explored as user interactions arrive. To leverage the initial latent vector estimates with the incoming interactions, we propose DynamicBPR – an algorithm for updating the new item latent vectors without retraining the CF model. Numerical evaluation reveals that DynamicBPR achieves similar accuracy as a CF model trained on all the ratings, using 70% less exploring users than conventional random exploration.

  • LPAdditive Co-Clustering with Social Influence for Recommendation by Xixi Du, Huafeng Liu and Liping Jing

    Recommender system is a popular tool to accurately and actively provide users with potentially interesting information. For capturing the users’ preferences and approximating the missing data, matrix completion and approximation are widely adopted. Except for the typical low-rank factorization-based methods, the additive co-clustering approach (ACCAMS) is recently proposed to succinctly approximate large-scale rating matrix. Although ACCAMS efficiently produces effective recommendation result, it still suffers from the cold-start problem. To address this issue, we propose a Social Influence Additive Co-Clustering method (SIACC) by making use of user-item rating data and user-user social relations.

    The main idea of SIACC is to extract the social influences from the social network, integrate them to additive co-clustering for effectively determining the user clusters and item clusters, minimize the loss error by backfitting the residuals of data approximation in the previous iteration, and finally improve the recommendation performance. In order to take advantage of social influence, we present a graph-regularized weighted-Fuzzy C-Means algorithm (gwFCM) to cluster users. gwFCM has ability to identify user groups from both rating and social information. Specifically, gwFCM makes sure that a pair of users have similar cluster membership if they have direct social relation (denoted as local social influence), and that the user with higher reputation (denoted as global social influence) plays a dominate role in clustering process. The reasonable user clusters obtained by gwFCM can benefit the item clustering, which will leverage the additive co-clustering processing and further improve the recommendation performance. A series of experiments on three real-world datasets have shown that SIACC outperforms the existing popular recommendation methods (PMF and ACCAMS) and social recommendation methods (SoReg, TrustMF, Locabal and SPF), especially on the cold-start users recommendation and running time.

  • LPFolding: Why Good Models Sometimes Make Spurious Recommendations by Doris Xin, Nicolas Mayoraz, Hubert Pham, Karthik Lakshmanan and John R. Anderson

    In recommender systems based on low rank factorization of a partially observed user-item matrix, a common phenomenon that plagues many otherwise effective models is the interleaving of good and spurious recommendations in the top-K results. A single incongruous recommendation can dramatically impact the perceived quality of a recommender system. In this work, we investigate folding, a major contributing factor to spurious recommendations. Folding refers to the unintentional overlap of disparate groups of users and items in the low-rank embedding vector space, induced by improper handling of missing data. We formally define a metric that quantifies the severity of folding in a trained system, to assist in diagnosing its potential to shock users with inappropriate recommendations. The folding metric complements existing information retrieval metrics that focus on the number of good recommendations and their ranks but ignore the impact of undesired recommendations. We motivate our definition of the folding metric on synthetic data and evaluate its effectiveness on both synthetic and real world datasets. We study the relationship between the folding metric and other characteristics of recommender systems and observe that optimizing for goodness metrics can lead to high folding and thus more spurious recommendations.

  • SPChemical Reactant Recommendation using a Network of Organic Chemistry by John Savage, Akihiro Kishimoto, Beat Buesser, Ernesto Diaz-Aviles and Carlos Alzate

    This paper presents recommendation algorithms for novel applications in chemistry. We focus on the task of recommending to the chemist candidate molecules (reactants) necessary to synthesize a given target molecule (product), which is an important step for the chemist to find a synthesis route to generate the product. We formulate this task as a link-prediction problem over a so-called Network of Organic Chemistry (NOC) that we have constructed from 8 million chemical reactions described in the US patent literature between 1976 and 2013. We leverage state-of-the-art factorization algorithms for recommender systems to solve this task. Our empirical evaluation demonstrates that Factorization Machines, trained with chemistry-specific knowledge, outperforms current methods based on similarity of chemical structures.

Back to Program

Diamond Supporter
 
Platinum Supporters
 
 
 
 
Gold Supporter
 
Silver Supporter
 
Special Supporters