2 2.1 SNS web Facebook Google+ SNS web SNS web HITS ANT(Auction Network Trust) web [4] SNS WEB PageRank HITS HITS web authorities, hubs Pagerank web S

Size: px

Start display at page:

Download "2 2.1 SNS web Facebook Google+ SNS web SNS web HITS ANT(Auction Network Trust) web [4] SNS WEB PageRank HITS HITS web authorities, hubs Pagerank web S"

えつまたかぎ
5 years ago
Views:

1 SNS Evaluation and Development reputation network for SNS user evaluation using realistic distance 1 3 1,2 Takanobu Otsuka 1 Takuya Yoshimura 3 Takayuki Ito 1,2 1 1 Center for Green Computing, Nagoya Institute of Technology 2 2 Department of Computer Science and Engineering, Graduate School of Engineering, Nagoya Institute of Technology 3 3 School of Techno-Business Administration, Graduate School of Engineering, Nagoya Institute of Technology Abstract: In recent years, SNS services such as Facebook, Google+, and Twitter are becoming very popular. In such services, many sources of information are posted and shared, although user rankings are hardly considered. In this paper, we consider an evaluation technique for web pages, such as HITS and PageRank, for SNS user evaluation applications and propose an algorithm using a user s real distance. We consider various parameters, including user distance, favorites, and the numbers of friends in SNSs in our evaluation technique. We propose a new reputation network to measure the reliability of SNS information. 1 Facebook SNS(Social networking Site) Twitter Facebook SNS [1][2] SEO (Search Engine Opitimize) SNS SNS Facebook otsuka.takanobu@nitech.ac.jp 1

2 2 2.1 SNS web Facebook Google+ SNS web SNS web HITS ANT(Auction Network Trust) web [4] SNS WEB PageRank HITS HITS web authorities, hubs Pagerank web SNS HITS PageRank SNS PageRank [3] web SNS SNS, [8] (Social Strength) [15][11] web SNS SNS [16] VCG [12] SNS web SEO Search Engene Optimization) web SNS Hack SNS km 2 HITS Pagerank 2

3 Distance SNS / 3.2 PageRank Google PageRank PageRank [10] P i PageRank r(p i ) P i PageRank B p P i P j P j P i PageRank r(p j ) PageRank n 1/n P i r(p i) P i,r 0 (P i) = 1/n PageRank PageRank Distance-HITS r k+1 (P i ) = { r k(p j ) + αd(p i, P j )} P j P j B pi d α 3.3 SNS SNS Facebook Facebook API Facebook SNS Facebook Facebook Graph.API / 3

4 Distance-Pagerank Facebook Java.csv GUI Pagerank Distance-Pagerank / Pagerank,Distance- Pagerank / 0 1.csv 1 Score スコアが最も高いユーザー IT ジャーナリスト User ID 計算パラメータの選択重みの設定 2: Pagerank ユーザー ID スコア 1: 4.2 Facebook 3 1. Pagerank 2. Distance-Pagerank 3. Distance-Pagerank / ID R 2 Pagerank Pagerank IT 2 IT 2 2 Distance-Pagerank 3 2 Pagerank 3 Distance-Pagerank 4 3 Distance-Pagerank 4

5 Score Score スコアの変動したユーザー群 User ID スコアの変動したユーザー群 User ID 3: Distance-Pagerank 4: Distance-Pagerank 4.3 Facebook / SNS FaceBook FacebookAPI 1 Pagerank HITS 5 Facebook web SNS,SNS SNS SNS SNS FacebookAPI 5

6 Facebook [1] Klout.inc, Discover and be recognized for how you influence the world., [2] Overtex Group, SNS Qrust., [3],,,,,, DEWS2008-Data Engineering Workshop,2008. [4],,, Auction Network Trust :,,Vol.J92-D, No [5] Taher H. Haveliwala, Efficient Computation of PageRank, 1999 Stanford Technical Report. [10] Taher H. Haveliwala, Efficient Computation of PageRank, 1999 Stanford Technical Report. [11] Bloch,F.and M.O.Jacsonjacson, The Formation of Networks with Transfers among Players,Journal of Economic Theory.2007 [12] Haoqi Zhang, Edith Law, Robert C. Miller, Krzysztof Z. Gajos, David C. Parkes, and Eric Horvitz, Human Computation Tasks with Global Constraints: A Case Study, Proceedings of the ACM Conference on Human Factors in Computing, [13] Z. Gy.Nongyi, H. Garcia-Molina, and J. Pedersen, Combating web spam with trust rank, Proceedings of the Thirtieth international conference on very large data bases, pp , VLDB Endowment, [14] S. Pandit, D.H. Chau, S. Wang, and C. Faloutsos, Netprobe: A fast and scalable system for fraud detection in online auction networks, Proceedings of the 16th international conference on World Wide Web (WWW 07), pp , [15] Matthew O.Jackson, SOCIAL AND ECO- NOMIC NETWORKS,Princeton University Press,2008. [16] Paul Adams, GROUPED:How small groups of friends are key to influence on the social web,new Riders,2012. [6] S. Brin and L. Page, The anatomy of a large-scale hyper textual web search engine, WWW7/Computer Networks, vol.30, no.1-7), pp , [7] L. Li, Y. Shang, and W. Zhang, Improvement of hits-based algorithms on web documents, Proceedings of WWW2002, pp , [8] Eric Gilbert and Karrie Karahalios, Predicting Tie Strength With Social Media, Proceedings of the 27th international conference on human factors in computing systems, [9] Josep M. Pujoi, Ramon Snguesa, and Jordi Delgado, Extracting reputation in multi Agent Systems by Means of Social Network Topology, Proceedings of the first international joint conference on Autonomous agents and multiagent systems, pp ,

2 JAWS web web Share = authorities ReShare = (hubs SNS i j i authorities j i i hubs 1 User i 情報が j によってシェアされる (authorities) j の情報をシェアする (hu

2 JAWS web web Share = authorities ReShare = (hubs SNS i j i authorities j i i hubs 1 User i 情報が j によってシェアされる (authorities) j の情報をシェアする (hu JAWS2012 JAWS2012 SNS Evaluation and Development reputation network for SNS user evaluation using realistic distance 1 Takanobu Otsuka Takuya Yoshimura Takayuki Ito Center for Green Computing, Department