Incremental Matrix Co-factorization for Recommender Systems with Implicit Feedback

• Type: Article in International Conference Proceedings Book
• Date: 2018

Authors

Anyosa, SC (Author) Other; Vinagre, J (Author) Other; Jorge, AM (Author) FCUP

Conference proceedings International

Title: Companion of the The Web Conference 2018 on The Web Conference 2018, WWW 2018, Lyon , France, April 23-27, 2018

Pages: 1413-1418

Indexing

Scopus - 13 Citations

Other information

Authenticus ID: P-00N-X4P

DOI: 10.1145/3184558.3191585

Abstract (EN): Recommender systems try to predict which items a user will prefer. Traditional models for recommendation only take into account the user-item interaction, usually expressed by explicit ratings. However, in these days, web services continuously generate auxiliary data from users and items that can be incorporated into the recommendation model to improve recommendations. In this work, we propose an incremental Matrix Co-factorization model with implicit user feedback, considering a real-world data-stream scenario. This model can be seen as an extension of the conventional Matrix Factorization that includes additional dimensions to be decomposed in the common latent factor space. We test our proposal against a baseline algorithm that relies exclusively on interaction data, using prequential evaluation. Our experimental results show a significant improvement in the accuracy of recommendations, after incorporating an additional dimension in three music domain datasets. © 2018 IW3C2 (International World Wide Web Conference Committee), published under Creative Commons CC BY 4.0 License.

Language: English

Type (Professor's evaluation): Scientific