Robust Cost-Sensitive Learning for Recommendation with Implicit Feedback

نویسندگان

Peng Yang

Peilin Zhao

Xin Gao

Yong Liu

چکیده

Recommendation is the task of improving customer experience through personalized recommendation based on users’ past feedback. In this paper, we investigate the most common scenario: the user-item (U-I) matrix of implicit feedback (e.g. clicks, views, purchases). Even though many recommendation approaches are designed based on implicit feedback, they attempt to project the U-I matrix into a low-rank latent space, which is a strict restriction that rarely holds in practice. In addition, although misclassification costs from imbalanced classes are significantly different, few methods take the cost of classification error into account. To address aforementioned issues, we propose a robust framework by decomposing the U-I matrix into two components: (1) a low-rank matrix that captures the common preference among multiple similar users, and (2) a sparse matrix that detects the user-specific preference of individuals. To minimize the asymmetric cost of error from different classes, a cost-sensitive learning model is embedded into the framework. Specifically, this model exploits different costs in the loss function for the observed and unobserved instances. We show that the resulting non-smooth convex objective can be optimized efficiently by an accelerated projected gradient method with closed-form solutions. Morever, the proposed algorithm can be scaled up to large-sized datasets after a relaxation. The theoretical result shows that even with a small fraction of 1’s in the U-I matrix M ∈ R, the cost-sensitive error of the proposed model is upper bounded by O( α √ mn ), where α is a bias over imbalanced classes. Finally, empirical experiments are extensively carried out to evaluate the effectiveness of our proposed algorithm. Encouraging experimental results show that our algorithm outperforms several state-of-the-art algorithms on benchmark recommendation datasets.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Cold-Start Web Service Recommendation Using Implicit Feedback

Service recommendation becomes an increasingly important issue when more and more Web services are published on the Internet. Existing Web service recommendation approaches based on collaborative filtering (CF) seldom consider recommending services based on users’ ratings on services since such kind of explicit feedback is difficult to collect. In addition, the new user cold-start problem is al...

متن کامل

The Effect of Written Corrective Feedback on the Accuracy of Output Task and Learning of Target Form

The effect of error feedback on the accuracy of output task types such as editing task, text reconstruction task, picture cued writing task, and dictogloss task, has not been clearly explored. Following arguments concerning that the combination of both corrective feedback and output makes it difficult to determine whether their effects were in combination or alone, the purpose of the present st...

متن کامل

Fast Collaborative Filtering from Implicit Feedback with Provable Guarantees

Building recommendation algorithm is one of the most challenging tasks in Machine Learning. Although most of the recommendation systems are built on explicit feedback available from the users in terms of rating or text, a majority of the applications do not receive such feedback. Here we consider the recommendation task where the only available data is the records of user-item interaction over ...

متن کامل

A Boosting Algorithm for Item Recommendation with Implicit Feedback

Many recommendation tasks are formulated as top-N item recommendation problems based on users’ implicit feedback instead of explicit feedback. Here explicit feedback refers to users’ ratings to items while implicit feedback is derived from users’ interactions with items, e.g., number of times a user plays a song. In this paper, we propose a boosting algorithm named AdaBPR (Adaptive Boosting Per...

متن کامل