Abstract：Location-based service recommendation system can provide users with more effective personalized services.The existing recommendation system mainly adopts differential privacy and k-anonymity for privacy protection,but the problems of poor service quality and low security coefficient are very serious.For the reason,we proposed a track privacy protection service recommendation model,which is based on the pattern of preference perception,and firstly adopts the mix-zone combined clustering method.When requesting service,the similarity of users in mix-zone is quantified through clustering to protect users′personal information.Secondly,the preference perception algorithm based on differential privacy(PPBP)is adopted to protect the recommendation results,and the privacy risk assessment is carried out according to the user′s preference.By adding noise of different sizes to the assessment results,the differential privacy budget is reasonably allocated to achieve the improvement of service.
李晓会,陈潮阳,白雨靓,张兴. 一种轨迹隐私保护服务推荐模型研究[J]. 小型微型计算机系统, 2021, 42(5): 990-995.
LI Xiao-hui,CHEN Chao-yang,BAI Yu-liang,ZHANG Xing. Research on a Service Recommendation Model for Trajectory Privacy Protection. Journal of Chinese Computer Systems, 2021, 42(5): 990-995.
2021, Vol. 42 
