Ensuring Security and Privacy Preservation for the Publication of Rating Datasets

Recommender systems are proposed to recommend the suitable artifact(s) to the target user. They are applied in several real-life systems such as Google, Facebook, Twitter, eBay, Amazon, PlayStore’s Android, and AppStore’s Apple. Generally, they are based on ratting datasets. Aside from recommender systems, the ratting datasets can also be shared with the data analyst. However, they have serious issues that must be considered when they are utilized, e.g., privacy violation issues. To address privacy violation issues in rating datasets, a privacy preservation model, ( \(l^{p_1}, \dots ,l^{p_n}\) $l^{p_1},\cdots ,l^{p_n}$ )-Privacy, is proposed. Although this privacy preservation model can address privacy violation issues in ratting datasets, it is highly complex and less effective. To rid these vulnerabilities of ( \(l^{p_1}, \dots ,l^{p_n}\) $l^{p_1},\cdots ,l^{p_n}$ )-Privacy, a new privacy preservation model for rating datasets is proposed in this work. With the proposed model, aside from privacy preservation issues, the complexity and the data utility are maintained as much as possible. Furthermore, the proposed model’s effectiveness and efficiency are evaluated by extensive experiments. From the experimental results, they show that the proposed model is an effective and efficient privacy preservation model for ratting datasets.