13 4 5 6 7 Q1 8 aCollege of Computer Science, Zhejiang University, 310 9 bCollege of Information, Zhejiang University of Financ 10 cDepartment of Computing, Macquarie University, Syd 11 d Stanford University, Stanford, CA 94305, USA 12 eDalarna University, SwedenQ2 13 1 5 a r t i c l e i n f o 16 17 18 19 20 21 Q4 22 23 24 25 26 27 2 8 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 en their te 54 ng the sell 55 the goods; hence, they experience a high degree of uncertainty . Meanwhile, social media tools such as social netw 56 services (SNS) enable people to share their opinions regarding a product and transaction [39,53]. To this end, an inc 57 number of e-commerce industries have adopted SNS to encourage user interactions, including eBay.com, Amazon.com, and http://dx.doi.org/10.1016/j.ins.2014.09.036 0020-0255/ 2014 Published by Elsevier Inc. ⇑ CorrespondingQ3 author at: Stanford University, Stanford, CA 94305, USA. Tel.: +1 650 319 5822, +86 13588029069; fax: +86 571 87951453.
E-mail addresses: email@example.com, firstname.lastname@example.org (X.-L. Zheng).
Information Sciences xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Information Sciences journal homepage: www.elsevier .com/locate / ins
INS 11139 No. of Pages 22, Model 3G 7 October 2014
Buyers in e-commerce can neither physically examine a product nor verify the reliability of the seller giv and spatial separation from sellers . In such a context, buyers usually have limited information regardiPlease cite this article in press as: S.-R. Yan et al., A graph-based comprehensive reputation model: Exploiting the social context of o to enhance trust in social commerce, Inform. Sci. (2014), http://dx.doi.org/10.1016/j.ins.2014.09.036mporal ers and orking reasingfactors to mitigate the subjectivity of opinions and the dynamics of behaviors. Furthermore, we enhance the model by developing a novel deception filtering approach to discard ‘‘bad-mouthing’’ opinions and by exploiting a personalized direct distrust (risk) metric to identify malicious providers. Experimental results show that the proposed reputation model can outperform other trust and reputation models in most cases. 2014 Published by Elsevier Inc.Article history:
Received 26 January 2014
Received in revised form 31 August 2014
Accepted 24 September 2014
Available online xxxx
Trust and reputation
Theory of reasoned action
Risk tolerance027 Hangzhou, China e and Economics, 310018 Hangzhou, China ney, NSW 2109, Australia a b s t r a c t
Social commerce is a promising new paradigm of e-commerce. Given the open and dynamic nature of social media infrastructure, the governance structures of social commerce are usually realized through reputation mechanisms. However, the existing approaches to the prediction of trust in future interactions are based on personal observations and/or publicly shared information in social commerce application. As a result, the indications are unreliable and biased because of limited first-hand information and stakeholder manipulation for personal strategic interests. Methods that extract trust values from social links among users can improve the performance of reputation mechanisms. Nonetheless, these links may not always be available and are typically sparse in social commerce, especially for new users. Thus, this study proposes a new graph-based comprehensive reputation model to build trust by fully exploiting the social context of opinions based on the activities and relationship networks of opinion contributors. The proposed model incorporates the behavioral activities and social relationship reputations of users to combat the scarcity of first-hand information and identifies a set of critical trustA graph-based comprehensive reputation model: Exploiting the social context of opinions to enhance trust in social commerce
Su-Rong Yan a,b, Xiao-Lin Zheng a,d,⇑, Yan Wang c, William Wei Song e, Wen-Yu Zhang bpinions 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 2 S.-R. YanQ1 et al. / Information Sciences xxx (2014) xxx–xxx
INS 11139 No. of Pages 22, Model 3G 7 October 2014
Q1Taobao.com. Product review sites such as Epinions.com utilize the same tools, which are part of the larger emerging phenomenon of social commerce [3,30,41,50] wherein the business activities of companies are supported by the voluntary effort of external partners . Therefore, the use of such social media generates new revenue opportunities for marketers and businesses in online shopping while providing consumers with product information and advice. These parties obtain both economic and social rewards for sharing .
With the aid of social media tools, each user within a community should ideally have the same communicative potential.
However, interested parties or stakeholders can easily manipulate online reviews for their strategic interests given the open and dynamic nature of social media infrastructure. Consumers also have external incentives to misreport and thus misrepresent the reviews available to other users [2,12]. Consequently, potential buyers discount actual reviews heavily as a result of the veracity of reviews questioned under deceptive environments.
Consumers are driven to value the decisions and opinions of social relationship members in product purchasing as per basic behavioral psychology [28,33]. However, new consumers usually have limited or virtually no direct interaction or relationship with other consumers in the context of social commerce because of the strong community structure in social networks . This observation implies that new entrants often serve only the small communities (cliques) of direct acquaintances. A review is the subjective perspective of a consumer regarding his/her experiences in community activities and may merely represent his/her individual preference and opinion. Furthermore, trusted friends may not have similar tastes. Consequently, users who increasingly rely solely on their acquaintance communities for information readily encounter collective community bias . In this case, the objectivity of reviews is not fully guaranteed under subjective environments.