earticle

논문검색

Is Trust Transitive and Composable in Social Networks?

원문정보

Hee Seok Song

피인용수 : 0(자료제공 : 네이버학술정보)

초록

영어

Recently, the topic of predicting interpersonal trust in online social networks is receiving considerable attention, because trust plays a critical role in controlling the spread of distorted information and vicious rumors, as well as reducing uncertainties and risk from unreliable users in social networks. Several trust prediction models have been developed on the basis of transitivity and composability properties of trust; however, it is hard to find empirical studies on whether and how transitivity and composability properties of trust are operated in real online social networks. This study aims to predict interpersonal trust between two unknown users in social networks and verify the proposition on whether and how transitivity and composability of trust are operated in social networks. For this purpose, we chose three social network sites called FilmTrust, Advogato, and Epinion, which contain explicit trust information by their users, and we empirically investigated the proposition. Experimental results showed that trust can be propagated farther and farther along the trust link; however, when path distance becomes distant, the accuracy of trust prediction lowers because noise is activated in the process of trust propagation. Also, the composability property of trust is operated as we expected in real social networks. However, contrary to our expectations, when the path is synthesized more during the trust prediction, the reliability of predicted trust did not tend to increase gradually.

목차

Abstract
 1. Introduction
 2. Related Studies
  2.1 Concept of Trust
  2.2 Propagation-Based Trust Prediction Studies
 3. Building Trust Prediction Models
  3.1 Trust Computation Strategy for the Experiment of Transitivity on Trust
  3.2 Trust Computation Strategy for the Experiment of Composability on Trust
 4. Experiments
  4.1 Data Set
  4.2 Experimental Results
 5. Conclusion
 References

저자정보

  • Hee Seok Song MIS Department in Hannam University

참고문헌

자료제공 : 네이버학술정보

    함께 이용한 논문

      ※ 기관로그인 시 무료 이용이 가능합니다.

      • 4,800원

      0개의 논문이 장바구니에 담겼습니다.