Research on Privacy Protection in Collaborative Filtering Recommendation System under the Needs of Information Sharing
Traditional recommended systems perform recommendation work on existing isolated database;howerver,information sharing for recommendation purposes are becoming increasingly popular with the evolution of the Internet for supplying better recommendation services for ecustomers. Due to privacy reasons while enjoying information sharing data we needs to consider privacy protection for data holders. Here,based on the diversity of the data source format,we standardize and format the source data through the ETL (Extract Transform Load) technology and then use k-anonymity model to protect each data holders privacy to integrate various shared data to a unified data model,thereby providing a unified platform which is based on the classic service algorithm for data sharing. We test and verify the solution is available for its privacy protection of data providers and still provide accurate recommendation results under the recommendation system under Collaborative filtering algorithms.
recommended systems information sharing ETL privacy protection k-anonymity Collaborative filtering
Han Lu Sun Lei
Information of Science and Technology Institute,East China Normal University,Shanghai 200241,China Information of Science and Technology Institute,East China Normalniversity,Shanghai 200241,China
国际会议
南宁
英文
192-195
2010-12-10（万方平台首次上网日期，不代表论文的发表时间）