Trust Level Of Users Research Articles

소셜 네트워크 서비스를 위한 키 분배와 사용자 평판을 이용한 접근 제어 메커니즘

최근 인터넷이 웹 2.0 시대에 들어서면서부터 페이스북, 트위터, 유튜브 등과 같은 많은 소셜 네트워크 서비스들이 생겨났다. 이를 통해 사용자들은 온라인 상에서 다른 사용자들과 관계를 맺고 그룹에 가입할 수 있으며 타인의 생활을 실시간적으로 확인할 수 있다. 하지만 무방비한 온라인 상의 정보 노출은 악의적인 사용자들로 하여금 쉽게 타인의 개인정보를 수집하고 이용할 수 있게 만들 수 있다. 이에 본 논문에서는 개인 정보의 남용을 방지하고 권한이 부족한 사용자들이 타인의 개인 정보에 접근하는 것을 제어하기 위한 접근 제어 메커니즘을 제안한다. 본 논문에서는 제안하는 접근 제어 메커니즘은 마스터키를 핵심키와 부분키로 구분하고, 요청자의 평판도에 따라 접근 권한을 제한하는 특징을 가지고 있다. 이러한 제안 접근 제어 메커니즘의 성능 분석을 위해 기존 소셜 네트워크에서 고려하는 정보 보호 메커니즘과의 성능을 비교 분석한 결과 복잡도와 계산 시간 소모량 그리고 키 관리의 안전성 측면에서 성능 향상이 가능함을 확인할 수 있었다. Recently, as Internet enters WEB 2.0, many social network services through such as Facebook, Twitter and Youtube appeared. In these social network sites, users can easily make friends, join groups and access others personal information. Therefore, a malicious user can easily gather information of others. In order to protect user's personal information from the unauthenticated users, we propose privacy protection mechanism based on key assignment and user's trust level. A master-key is generated for each users and is segmented into a core-key and several sub-key. The master-key stores at the information owner's side and the sub-key will be distributed to requestor according to the relation and trust level. At last, in order to proof the efficiency, the performance of our proposed mechanism is compared with those of existing mechanisms.

Dynamic Trust for Secure EShopping in Semantic Web

present scenario, electronic shopping has become a trend which is easy and time-saving. Such convenient services on web and the products provided by sellers are trustworthy or not are very important for buyers to know. Seller can even increase their product rating by acting as a buyer or can decrease the rating of his competitor's product. Also a buyer can also give the wrong feedback. So it is difficult to trust such e- products and user's feedback. In this paper, a dynamic trustworthy framework rectifies lack of static trust in semantic web scenario on the basis of the trust level of users (Buyers and Sellers). To achieve this, three agents are employed (namely Process Agent (PA), Detection Agent (DA) and Trust Calculation Agent (TCA)) for calculating dynamic trust in Semantic Web. These intelligent agents collectively restrict any imposter from increasing the fake trust value and hence furnish a genuine trust value dynamically.

User security management based on trust constraint

It is very difficult to identify the malicious acts of users and to control the future acts of users for the existing static certificate-based user management in an open grid environment because of the dynamics and uncertainty of user behavior. To address the problem, a method for user security management based on trust constraints was proposed. In this method, according to the result of trust computation, the trust black-list and white-list were established by monitoring the change of trust level of users, and trust policies that stated the relationship between user trust and asset value were set. The testing results show that, as a dynamic trust constraint of access control mechanism, the trust black-list, white-list and trust policies strengthen the constraints on the system authorization, and effectively control the user behaviors by associating user's capacity accessing to resources with his or her history behaviors, which improves system security.