2510712017 _Difusing mut-aspects of local and global social tust for personalizing rust enhanced recommender system - IEEE Xplore Document IEEE.org | IEEE Xplore Dial Library |IEEE-SA| IEEE Spactrum | More Stos Cart (0) | Create Account | Personal Sign In Institutional Sign In Diffusing multi-aspects of local and global social trust for Related Articles: personalizing trust enhanced recommender system ev no guaran rider es fora SS, Flv Sctestyn pose onet ota 2 47, Sentniteumar; R, Ponnusamy View AIL Authors aa Abetract | Autiors Figures References | Ciaions | Keywords Metres Madin Abstract: Recommender systoms are an effective solution to the information overload problem, especial in tho World Wide Wel where we gather vast Information from anonymous people around the worl, Trust enhanced recommender system tobe promising to overcome the coldstart and sparsity challenges of tradtional recommender system as well as improving the accuracy ofthe recommendations. This arse a research focus about blending ofthe trust information and trust level prediction tothe recommendation framework. From the past decade, numerous researches ‘wore done to adapt one social notwork trust (simply social trust) for many web applications, including o-commoreo, P2P networks, mul-agent systems, recommendation systems, and service-oriented computing. Usually, online socal ust prediction can be based on two mecnanisms to acquire trust value: evaluating tusioe on bass of tusterustrs neighbor tust experience information (local ust), otherwise evaluating trustee on the basis ofthe whole social network tust experience information as reputation (global ust). Here, we leverage socal science teas to develop ‘the trust models that enable the study of online socal trust evokton, In his paper, we propose a matrix factorization based trust enhanced recommendation system which properly incorporates both local rust and global rust with ifusion of the social rust mult-aspects to improve the {ually of recommendations for miigaing tho data sparsity andthe col-start issues. Through experiments on the Epinions dataset, wo show that ‘ur model outperforms its standard trust enhanced counterparts with respect to accuracy on recommender systems, Published in: Advanced Computing and Communication Systems ((CACCS), 2016 Sr International Conference on Date of Conference: 22:28 Jan. 2016 INSPEC Accession Number: 16375351 Date Added to IEEE Xplore: 10 October 2018, Dor: 10.1108"CACes.2016.7586987 ISBN Information: Publisher: IEEE Electronic ISBN: 978-1-4672-9206-8 Print on Demand(Pob) ISBN: 978-1-4673-9207-5, Conference Location: Coknbatoce, nda : Contents L introduction “Trusts one source of information that has been widely adopted to personalize web applications, including e-commerce, P2P networks, mull-agent systems, recommendation systems, and service veieweciiors rented computing. However, tus information is usualy very sparse or unavailable for most online Fall Text systems [1] [19], [111 [2]. Onine social network based trust are becoming greater significance and Wen tron practicaty availabe with friendship links and fllowereelatlonships which is utlized forthe Abstract predicton of tust (inference and propagation) (12, (3), [19} [5]. [4]. Mosty, many e-commerce and ma consumer review websites provide both rviews of products and a social natwork structure among Authors the reviewers for analyzing usor' taste, These wobslos gives an important rle for combining users’ product reviews and ther online Socal network structure for predicting their socal intuence Figures hp:tieeexploeioee.orpldocument7586387) 182510712017 _Difusing mut-aspects of local and global social tust for personalizing rust enhanced recommender system - IEEE Xplore Document 'n buying ne proauc or sernce (21) nus, tization oF onine social eworK Information anNgs ‘new opportune to personalize ust enhanced recommendations, References Read document tations Keywords Keywords IEEE Keywords Back to Top Recommender systems, Social network services, Predictive models, Computational modeling, Collaboration, Communication systems, Sparse matrices INSPEC: Controled indexing rusted computing, matrx decomposition, recommender systems, social networking (orine) INSPEC: Non-Controlled indexing ‘epinions data set, global socal rust, local social rust, personalizing trust enhanced recommender system, information overlad problem, Worl! Wide Web, tust level prediction, e-commerce, P2P networks, mulagent systems, service-oriented computing, online socal network trust prediction, social science theory, mati factorization based trust enhanced recommendation system, online social rust evolution, socal rust mutiaspectdfusion, data sparsity, ‘Author Keywords. Matrix Factorization, Social Trust, Mult-aspects, Recommender System Authors KT. Senthikumar R&D Department, Sharathiar Universily, Coimbatore, India R. Ponnusamy Ral Gandhi Collage of Engineering, Chennai, India Related Articles Joint ajectory generation for redundant robots T.C. Hsia: ZY. Guo Reconfiguration of resources in middleware HA DurancLimon;G.S. Blair ‘Schadulabilty in model-based sofware development for dstrbuted real-tme systems SS. Yau; Xigoyong Zhou Performance evahation ofa probabilistic replica selection algorithm SS. Krishnamurthy: WH. Sanders: M. Cuker Distributed objoct-oronted reattime simulation of the muticast protocol RFRM Ys, Hong ‘Schedulabiliy analysis ofan event-based reattime protocol ramework 4J- Rodrigues J. Ventura L. Rocrigues Wavelet based lossy compression of barotropic turbulence simulation data JP. Wilson ‘Somi-gerote matt tansforms (SDD) for mage and video compression hpilieeexplore joes. orgldocument7586387! 282510712017 _Difusing mut-aspects of local and global social tust fr personalizing rust enhanced recommender system - IEEE Xplore Document SUI approach fo programming of TMO frames, KH. Kim; Seok-Joong Kang: Yuging Li Implementation of a TMO-based real-time airplane landing simulator on a dstibuted computing environment Min-Gu Lee; Sunggu Lee IEEE Account Purchase Details Profile Information Need Help? » Change Usemame Password » Payment Opens Communists Preferences ‘US & Canada: +1 209.678 4992 Update Ades » Orde story » Profession and Easton Wortdwide: +4 722961 0060, View Purchased Documents = Technica terest Contact & Suppo Abt IEEE Xplre| Conac i | Hol |Terms so Nandaorminaton Pele | Stomap| Privacy & Opting Out Cokie htptieeexploreioes.orgldocument7586387) a8