2014년 컴퓨터종합학술대회
Current Result Document :
| 한글제목(Korean Title) |
Trust Prediction Using Social Relations |
| 영문제목(English Title) |
Trust Prediction Using Social Relations |
| 저자(Author) |
Tae-wan Noh
Seo-hyang Kim
Nopphon Keerativoranan
Chong-kwon Kim
|
| 원문수록처(Citation) |
VOL 41 NO. 01 PP. 1145 ~ 1146 (2014. 06) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
trust prediction in social network becomes important issue since the increasing of online users who spends and provide services. Either similarity or social relation is used to predict trust relation among users. Our idea show that both can be used simultaneously as method to predict incorporated with low-rank matrix factorization. Our experiment uses Epinions's movie rating and review rating as dataset comparing new method and random method, generated Trust relation randomly, by prediction accuracy. The result shown 1) new method has higher accuracy while the 2) prediction accuracy has constant instead of increasing when number of testing subject is larger.
|
| 키워드(Keyword) |
|
| 파일첨부 |
PDF 다운로드
|
