Research And Implementation Of Social Network Privacy Disclosure Risk Assessment Technology

With the rapid development of online social networks and the growth of user scale,the online social networks also bring the risk of privacy information leakage to the vast user group.Therefore,how to protect personal information on social networks has become an important issue on the development of social networks.Privacy disclosure risk assessment can effectively detect the privacy leakage status of social networks,which is of great significance to users themselves,social network platforms and social development.At present,research on privacy leakage risk assessment mainly focuses on the user’s perspective,and based on the attribute information and privacy preference settings provided by the user.A small amount of research considers the network environment and behavioral characteristics of the user.The comprehensiveness of privacy risk assessment is lacking.In response to the issue,this paper systematically investigates and analyzes the privacy leakage pathways in social networks,and establishes a comprehensive social network privacy leakage risk assessment model from the perspectives of users and platforms.This model not only considers the risks of users’ personal information disclosure,but also considers the impact of the platform on user privacy leakage,able to comprehensively assess the risk of privacy leakage on social networks.The specific research content is as follows:(1)The method of evaluating the privacy disclosure risk of social networks from the perspective of users themselves is studied.At present,most of the research is based on the profile information provided by users.This paper considers both user profile information and text content,which can more fully consider the user’s privacy disclosure path.When quantifying privacy based on user profile,the impact of data granularity on privacy disclosure is considered,and the privacy disclosure risk score of user profile is calculated in combination with sensitivity and visibility to make the evaluation result more accurate.When quantifying privacy based on text content,first use clustering algorithm to complete the construction of text privacy system,then use deep learning technology to identify the privacy entities in the text,and then quantize the privacy in the text content based on the information entropy theory.Finally,combined with user profile and text content privacy quantification,the user’s comprehensive privacy disclosure risk assessment quantification value is obtained.(2)The method of evaluating the privacy disclosure risk of social networks from the perspective of platform is studied,a risk assessment method of privacy leakage based on cloud model and improved evidence theory is proposed.Through the analysis of relevant documents,laws and regulations,the risk assessment index system is constructed from three aspects:platform vulnerability risk,network environment risk and external threat risk.Using cloud model to quantify indicators can solve the problem of fuzziness of expert evaluation to a certain extent.The use of evidence conflict correction method based on conflict coefficient and probability function and evidence synthesis algorithm based on matrix analysis can solve the problem of evidence conflict to a certain extent and improve the efficiency of risk assessment.(3)This paper takes the Weibo platform as the research object,through collecting user data and expert evaluation information,the applicability and effectiveness of the method in this paper is proved by case study,which can provide a new idea and method for privacy disclosure risk assessment of social networks.