Matrix Factorization for cloud service recommendation based on social trust

Abstract

Recommending trustworthy cloud services is essential to establishing credibility and ensuring better user decision-making based on their specific needs. Traditional recommendation approaches based on collaborative filtering face some challenges, including data sparsity issues. In this paper, a social trust-based recommendation approach using matrix factorization is proposed to improve recommendation accuracy and address the limitations imposed by data sparsity in recommender systems. First, the level of trustworthiness of users is inferred from their interactions on social networks. Then, the social trust model is integrated with the matrix factorization technique to generate reliable recommendations for users. The results of experiments conducted on the Epinions and WSdream datasets demonstrate that social trust significantly improves recommendation accuracy compared to state-of-the-art recommendation systems that do not take trust into account.