Foundations of Trust: Contextualising Trust in Social Clouds

Foundations of Trust: Contextualising Trust inSocial Clouds

Simon Caton,∗ Christoph Dukat,† Tilo Grenz,† Christian Haas,∗ Michaela Pfadenhauer,† and Christof Weinhardt∗∗Karlsruhe Services Research Institute / †Institute for Sociology,

Karlsruhe Institute of Technology, Germany{simon.caton, christoph.dukat, tilo.grenz, ch.haas, michaela.pfadenhauer, weinhardt}@kit.edu

Abstract—In this paper, we lay the foundations for a contextu-alisation of trust, the role it plays, and its different layers withinthe context of a novel paradigm: Social Cloud Computing. In aSocial Cloud, trust plays a vital role as a collaboration enabler.However, trust is not trivial to define, observe, represent andanalyse as precursors to understand exactly what role it plays inthe enablement of collaboration. We do this through the definitionof structure of a Social Cloud as a sequence of social and cognitiveprocesses. We then survey research from the domains of computerscience, economics and sociology that consider trust in onlinecommunities and exchange scenarios to illustrate the complexityof modelling trust in our scenario. Finally, we define trust withinthe context of a Social Cloud and identify the core componentsof trust to facilitate its understanding.

I. INTRODUCTION

The ever increasing pervasiveness of Social Network Plat-forms has profoundly changed the way that we communicateand interact today. They allow us to interact within virtualplatforms, establish and engage in virtual communities, aswell as represent, document and explore our inter-personalrelationships in a digital manner. Their widespread uptake hasmeant that our electronic relationships are beginning to be in-tertwined with their real world counterparts, and in some casesare indistinguishable. Consequently, a large portion of socialinteraction now takes place via social network platforms.

In parallel, the rise of distributed computing and serviceparadigms like Cloud Computing have made the acquisitionand consumption of Internet services more common. Yet, stud-ies have shown that issues such as trust, security, anonymityand sometimes reliability are significant obstacles for theseparadigms. As an alternative, we defined the concept of aSocial Cloud in [1], [2] to build upon the today’s prolificuse of social network platforms, users’ increasing technicaladeptness thanks to web and Internet science, and that users’locally available capabilities and resources are dramaticallyincreasing. The latter means that many Internet users haveresource endowments that studies have shown to be 60-95%idle [3]–[5]. The volunteering computing paradigm has alsodemonstrated that users will make idle resources availablefor “good uses” either altruistically or without a directlyexpected utility. Consider that an average Facebook user has190 friends,1 and our vision of a Social Cloud: a resource

1http://tinyurl.com/fb-anatomy – last accessed May 2012

and service sharing framework that utilizes relationships es-tablished between members of a social network [2] emerges.

Social Clouds provide an environment in which (new)provisioning and sharing scenarios can be established basedupon implicit levels of trust that transcend from the inter-personal relationships digitally encoded within social networkplatforms. The vision of a Social Cloud is motivated bythe need of individuals or groups for specific resources orcapabilities that could be made available by connected peersand peer groups. In simple words, Social Clouds use socialnetworks as mechanisms for efficient collaboration, as usersleverage their existing networks to share capabilities andresources. Resources are not necessarily only computationalresources, but can be any electronically exchangeable service,including human resources, skills and capabilities.

This definition contains a key assumption: using trust asa vehicle for exchange implies that leveraging existing rela-tionships are better or more efficient than anonymous-basedexchanges. It also does not reflect the type of trusted relation-ship needed for a successful exchange. The concept of trusttherefore requires an analytical understanding as a concept ofsocial action and context. Therefore, in this paper we examinethe context of trust and its key aspects for online collaboration.We address the challenge of defining and understanding trust,such that it can be leveraged as an exchange enabler for aSocial Cloud, specified and investigated experimentally.

This paper is organised as follows: section II, describesthe structure of a Social Cloud. Section III, surveys trust incollaboration contexts. Section IV, discusses a trust definitionfor a Social Cloud. Section V, summarises the paper.

II. SOCIAL CLOUD PARTICIPATION AND INTERACTION

To help define the context of a Social Cloud and the rolesof trust, we briefly describe the structure of a Social Cloudas a sequence of social and cognitive processes (Fig. 1).We present three stages: Prior Expections, Social Interchangeand Completion. Their cyclic dependencies demonstrate theevolutionary nature of interactions within a Social Cloud, andhighlight how past interactions influence user expectations.

A. Prior Expectations

The ex-ante stage prior expectations captures why a userwould consider using or joining a Social Cloud, how they

Fig. 1. Social Cloud Exchange Structure

envisage their benefit of its use and the contribution they canmake. We identify two aspects that shape the expectations of auser: their motivation, and the existence of demand and supply.

Motivation is expressed via the following aspects: 1) ex-pected outcome of the social exchange, 2) social context and3) individual as well as network history. The outcome ofan exchange captures several elements of providing as wellas consuming resources. This includes aspects like a gain inutility, fulfillment of a goal, completion of a task, and feelingof inclusion or usefulness. These actions incur a sense ofbelonging or togetherness via participation in a Social Cloud.Yet we note, that specific incentive schemes (see [6]) may alsobe drivers of participation. The social context captures specificproperties of the Social Cloud and its users: relationshiptypes (e.g. family, close friend, acquaintance, colleague etc.);specific properties of the social graph (e.g. centrality, andconntectivity); and finally the implicit levels of trust betweena user and their friends. Trust plays a crucial role and containsseveral frames of reference: 1) trust as an intrinsic (subjective)attribute of an inter-personal social relation, i.e. trust as a basicfoundation of social actions; 2) trust in the competence of anindividual to be able to deliver a given resource or capability,i.e. a belief in the self-awareness and personal re-evaluationof self-efficacy undertaken by an individual as a precursor topartaking in a Social Cloud; 3) trust in an individual to deliver,i.e. keep their promises, adhere to any (informal) agreementsetc. The final aspect of user motivation is the observation ofinteraction history, which is described in the completion stage.

Supply is the circumstantial availability of a useful resourceor capability in a Social Cloud. In other words, excess drivendemand - where the need of an individual is satisfied by thecircumstantial offering of excess resources or capabilities bytheir peers. Demand is the observation of individual needs thatencourage users to provide to their peers, i.e. it is demand-induced social capital - where the need of one or more usersin the Social Cloud compels other users to make their excessresources or capabilities available as a form of social capital.

B. Social Interchange

Social Interchange entails the facilitatation of collaborationand exchange between socially-connected peers. Many meth-ods exist for the construction and representation of this stage,which we do not elaborate on here. Instead, we illustrate thecore components of a platform for collaborative social ex-

change. We differentiate between formal and informal aspects,where formal aspects define collaboration as an economicsystem and informal define the social and context specificstructures that help facilitate exchange as a socially-drivenprocess. In general, the core aspects of this stage are focusedaround: the identification and allocation of demand and supplyusing socio-economic mechanisms; messaging mechanisms(for communication and coordination), or other mediumsthat can be compared to social network platforms; and theprovisioning and delivery of exchanged artefact(s).

C. Completion

The ex-post stage Completion addresses the processes andactions that finalise an exchange, and contains three keycomponents: feedback, recommendation and archiving. Com-pletion provides an element of closure to a social exchangeand potentially results in a positive or negative change to therelationships of users, as completion does not imply success.

Feedback is the social dissemination of an exchange as anexercise in reflection and communication via two modes: 1)local feedback, i.e. between users (potentially privately), and2) public feedback via social channels like notifications, news-feeds, and Facebook’s Timeline. Feedback includes: the benefitto consumer, potentially a reward given to the provider or amessage of thanks. For a Social Cloud, feedback demonstratesusefulness, is a means to attract users, encourage participation,and demonstrates the existence of trust.

Recommendation is the suggestion of an action after apositive and negative exchange outcome, i.e. whether to rewardor sanction a user. This process is heavily dependent on thesocial context of a collaboration. In other scenarios, suchactions would be defined in service level agreements. Yet, thisis difficult to argue in the case of a Social Cloud, as rigidlyengineered forms of interaction completion cannot considerthe intricate details of a social context. It may also not bepossible to determine if an exchange was successful or not,without additional information supplied by users.

Interaction Archiving is the documentation of collabora-tion processes. Although a simple task, it is critical for theunderstanding of trust, its evolution between users and itssocial context. Archives act as repositories for determiningcollaboration performance and social cohesion (often the resultof interaction history) as a result of specific inter-individualexchanges based on positive expectation (trust).

III. SURVEY OF RESEARCH INTO TRUST

Trust in the context of a Social Cloud requires an in-terdisciplinary understanding. In this section, we present aninterdisciplinary survey as a basis for discussion. We selectComputer Science with a focus on the technical implementa-tion of trust; economics with a focus on how trust effects theeconomic behaviour of a system’s users, the system itself aswell as economic situations where trust plays a crucial role;and Sociology with a focus on the theoretical basis for theexplaination of context (e.g. history, setting, and expectations)and fundamental characterisation of (social) exchange.

A. Trust in Information and Computer Science

With the advent of electronic platforms where users caninteract with each other, such as e-commerce or P2P platforms,is has become necessary to define and implement computa-tional notions of trust. In order for interactions to occur inthese types of platforms, the exchange between users has to befacilitated by technical features that address and try to alleviatethe trust issue in such systems. Building on research from otherdisciplines, computer and information science has focused onapplying these principles to technical systems. Examples arereputation systems to create a certain level of trust (e.g. eBay,and P2P networks), sharing platforms targeted at a certainuser group (e.g. MyExperiment), algorithms to calculate trustscores between users, and approaches that leverage socialnetworks as a basic premise for collaborative exchange.

Today, investigations into adopting social network struc-tures for different types of collaboration are in their infancy.Key examples are: community and scientific portals likePolarGRID [7] and ASPEN [8]; social storage systems likeFriendstore [9], and omemo.com; the sharing of networksand infrastructure (e.g. fon.com), the distribution of insurancepolicies amongst social peers (friendsurance.de) and wheresocial networks emerge due via collaboration, e.g. [10], [11].

In such cases, the social network (and therefore “trusted”context) emerges artificially as a consequence of use or thesocial network platform is used to perform mundane taskslike user authentication, and messaging. In both cases, it isimpossible to try and understand the context of trust behindsuch approaches. Alternatively, the approach exists outside ofthe social network platform. In such cases, no considerationis made on how the digital relationship representation canbe leveraged detaching the notion of trust as a collaborationvechicle. The key exception here is Friendsurance, whichalthough it doesn’t explicitly harness the social graph, itleverages existing relationships between friends to enablefriend- and community-based insurance contracts. Here, smallinsurance claims are covered by friends or the community, andlarger ones are given to insurance company resulting in lowerinsurance fees. Trust in this context, however, is somewhatovershadowed by the system’s legal (i.e. contractual) context.

Three notable approaches to sharing resources in socialcontexts are MyExperiment.org [10], nanoHUB.org [11] andSocial Volunteer Computing [12]. MyExperiment is a platformwhere users can provide and work on workflows for scientificprocesses which facilitates the sharing and dissemination ofcommon workflows. nanoHUB lets users share (teaching andresearch) resources in the field of nanotechnology and usesa simple virtual currency as participation incentives to users.Social Volunteer Computing is intended as the next generationof Volunteer Computing where users have underlying socialrelationships. However, each of these concepts are specializedto certain types of resources, do not extend beyond their virtualcontext, or do not consider actual sharing due to a missingbilateral focus. Hence, they lack the setting to investigate therole of trust and leverage it in exchanges.

In the integration of trust in electronic systems, several rep-utation mechanisms that measure, represent or use trust scoreshave been proposed (e.g. [13]–[16]). Their aim is to calculatea measure of trust based on feedback of users, previous inter-actions, and other factors. Social Regret [14] is an augmentedreputation system that leverages social relationships and socialnetwork analysis. It considers private and forms of communityreputation (such as neighborhood reputation) and takes intoaccount the reliability of reputation information. EigenTrust[15] is a means to decrease the number of fake files in P2Psharing networks. Through the calculation of a global trustscore for each peer, and with the approach that users shouldselect their download peer based on this global trust value,the authors show that the number of fake/corrupt files in thenetwork significantly decreases. Trust in this context can beseen as selecting a download peer without receiving fake files.PeerTrust [16] aims to calculate the trustworthiness of peersin a P2P exchange scenario, taking into account aspects suchas feedback from other users and its credibility, number oftransactions and even the context of transactions. It definesa trust metric that aggregates the individual trust measuresto a single number. PowerTrust [13] leverages power-lawdistributions of user feedback and a trust network to improvethe accuracy and scalability of a reputation system. All ofthese systems focus on the measurement and implementationof trust scores, taking the existence and influence of trust asa given or don’t consider social contexts.

B. Trust in Economics

Where computer and information science focuses on thetechnical implementation of trust in (electronic) systems, eco-nomics focuses on how trust effects the behaviour of marketparticipants and what role trust plays in the functioning ofa market itself. Trust is seen as an important facilitator ofmarkets (see e.g. [17]), as there is the risk of opportunisticdefault in many market settings (i.e. sellers do not ship afterpayment, or buyers do not pay for goods). Or, as [18] putsit: “Virtually every commercial transaction has within itselfan element of trust, certainly any transaction conducted overa period of time. It can be plausibly argued that much of theeconomic backwardness in the world can be explained by thelack of mutual confidence.” Kozinets’ [19] research on “virtualtogetherness” showed that exchange and cooperation changeswith the developmental progression of individual memberparticipation and interpersonal exchange, which is one of thebasic assumptions of trust from a sociological perspective.

The economic notions of trust, however, mainly focus on theeffect of trust relationships upon economic outcomes. Amongthe first, the existence of Social Networks and their interplayon economic processes has been studied by Granovetter [17],[20] who assessed the effect of social networks on job in-formation distribution. He found that the dissemination ofjob information depends on the structure of the underlyingnetwork, and that the “weak” ties between almost separate,clustered groups play important roles in the dissemination ofinformation among these groups. More recently, [21] study the

influence of social links and trust on business transactions suchas auctions. In particular, they study an online auction platformwhich includes social network capabilities. They find thatusers actually make use of their social network for businesspurposes or transactions, which benefits the users via higheruser satisfaction rates. Furthermore, trust in pure businesspartners decays faster with the path length between two users,compared to trust between users in the social network.

The effect of trust on economic outcomes has also beensubject to many game theoretical studies, mainly in the formof the Trust game and the notion of reciprocity [22]. In theTrust game, which can be seen as a study of reciprocity itself,one player decides if and how much money to send to thesecond player, in which case the second player receives amultiple of the amount sent. In the next step, the secondplayer can appreciate this gesture by sending a percentageback in return. Although the standard self-interest hypothesispostulates an equilibrium where no money is sent to the secondplayer, a significant fraction of first players exhibit trust bysending money without knowing if they will get something inreturn. Based on these results, economic models of trust in thesense of reciprocal actions have been proposed in [23]. Yet,the notion of trust that is considered in these approaches doesnot consider specific relationships or social contexts.

Economists have also made efforts to incorporate trust (andits effects) in their models. For example, [24] describes trustfor economic cooperation, addressing how trust affects eco-nomic outcomes and discusses how trust can be measured. An-other approach is the consideration of social or other-regardingpreferences [23], [25], [26]. The aim of this area of economicsis the explanation of empirical and laboratory findings whereusers do not act according to traditional economic assumptionssuch as Rational Choice. For example, in a Trust game,economic theory under the assumption of fully rational actors,predicts that nothing is shared. In contrast, the significantamount share in experiments is an indication that people placesome sort of trust in their (unknown) counterpart. This fact isexplained by several authors through different explanations,where reciprocal behavior and conditional cooperation are theconcepts applied by most authors. The problem with someof the earlier experiments is that sometimes the conceptsof other-regarding preferences are intertwined with trust andreciprocity, and it cannot be clearly inferred from experimentalresults where one of the factors truly is significant. To alleviatethis issue, [27] ran experiments to specifically study the influ-ence of trust vs. reciprocity in settings where other-regardingpreferences play a role. The results show that all three aspectshave a significant effect on the behaviour of participants, i.e.that they are different but equally significant factors.

C. Trust in Sociology

In sociology the phenomenon “trust” is mainly consideredas a form of calculated risk taking corresponding to specificdecision situations in social interactions. This understandingof trust can be found especially in rational choice [28] andsystems theory [29]. Beyond this, trust is a crucial aspect

in the sociological discussion of modernization processes,where we can observe the diminishing of traditional “familiar”structures, changes in the relations of nearness and distanceas well as increasing differentiations (e.g. diversification ofexpert knowledge) and technological forthcomings. Due tosuch developments and their corresponding risks, an increasingdemand of trust, especially system trust, and the role of insti-tutionalization processes to facilitate trust are identified [30].However, an integrated socio-theoretical conceptualization thatcaptures the complete phenomenon of trust is missing [31].

From a phenomenologically oriented sociology of knowl-edge perspective, Endress [32] tried to fill the gap by offeringthree levels of trust to tackle the limited view on trust hecalled reflexive trust that describes a “cognitive modus andstrategic resource” and refers to the understanding of trustespecially in rational choice theory. Endress stressed theimportance of the aspect of “familiarity” (of knowledge) asa fundamental resource for trust. Based on this, he added twofurther levels of trust: “operational trust” and “habitual trust”which address the fundamental social theoretical prerequisiteof the existence and the development of trust and distrust basedon insights from ethnomethodological [33], phenomenological[34] and institutional [35] approaches that correspond to theactions of people in everyday life. “Operational trust” isdefined as a “constitutive mode of trust”: a non-thematicelemental precondition of human action, which represents animplicitly unquestioned standpoint in the life-world [34], or“background expectation” according to [35]. Although it ishardly operationable, this level has to be kept in mind as afundamental social theoretical dimension when dealing withthe constitution of trust. The notion “habitual trust” is a“pragmatically effective fundament of routine and the productof interaction”. Upon this basis, we follow Endress’ definitionof trust according to Lewis and Weigert [36]: Trust has to beconsidered as an (explicit or implicit) reciprocal orientationof at least two actors, which rests on the (explicit or implicit)common shared understanding of situation and which comes toexpression in resulting structural behaviours and actions [37].

Besides the theoretical foundation of the complex conceptof trust not much fundamental empirical research specificallyon trust within modern social network sites like Facebookhas been undertaken. Such networks have to be regardedas the context in which trust relationships are systematicallyembedded. Research on social networks mostly focuses on the(transforming) quality of friendship or traditional relationshiptypes for which trust is an implicit but fundamental category[38]. Relationships and relationship practices have to be un-derstood as socio-historical products that are interrelated withsocietal reflections, images and conventions on friendship [39].Today, a significant degree are inscribed in social networkinfrastructures [38]. In contradiction to the 1990s notions ofsocial isolation, early studies on Facebook already show strongempirical evidence that networking people tend to a socialhomogeneity [40] which even resulted in social processesof differentiation and separation of other groups [41]. Forthe individual concept of close relationships (in particular

friendship) a significant dynamism is described in later studieswhich is seen as a consequence of the shift from static rela-tionship concepts to the instable practices (e.g. status updates,information posts, and a continual observation of others etc.),which Turkle [42] calls friended instead of a friend.

For questions on trust in such networks and together withthe ongoing empirical findings of homogeneity, it has beendiscussed if the (theoretical) notion of trust as an interactionaland proved product has to be extended to a common goal ofmaintaining established networks [43]. It was Granovetter whoalready mentioned the impact of the quality of relationshipson trust [44] (see also [45]). According to Granovetter’s ideait is necessary to consider the strength of relationships [20],the specific relationship dimensions [46] and concepts ofinternet ties [47] as well as the different roles they play forexchange processes in social networks and the usage of socialcapital [48]. Behind this background in empirical research itis necessary to regard both the level of reflexive and habitualtrust as well as connect them with the dynamic qualities ofdifferent relationships.

D. Summary

We have surveyed many concepts of trust in computerand information science, economics and sociology. Yet, noneproves sufficient for a fully rounded trust model that is fittingfor a Social Cloud. Research from computer and informationscience focuses on the technical implementation of trust inelectronic systems, but lacks the theoretical depth to under-stand trust. Instead, (often uninformed) models of trust arearchitected as a basis for decision making. Economics studiesthe effect of trust on economic outcomes, taking aspects suchas the existence of trust as a given. This provides a premise tounderstand how trust can be leveraged under certain economicassumptions, but not how trust can be understood as a driverfor exchange. Sociology studies the fundamentals of trust andinvestigates aspects such as the evolution of trust over time,its provenance and provides an contextualised understandingof trust as a multi-facteted social process. However, this iswithout any consideration of the technical context of trust foran online community, or its inclusion in exchanges.

IV. IDENTIFYING TRUST FOR A SOCIAL CLOUD

The realisation of a Social Cloud requires the design andimplementation of a social middleware and trust models thatcombine the insights and overcome the shortcomings fromthree perspectives: computer science with a focus on theimplementation of trust models for a Social Cloud; economicswith a focus on an economic system for interpersonal ex-change via trusted relationships; and sociology as bringingthe fundamentals for the understanding of trust. This re-quires the underlying and shared understanding of trust asbased on reciprocal implicit/explicit expectations accordingto different roles (consumer, provider) in diverse exchangecontexts between social peers. Under consideration of a sharedand interdisciplinary understanding of trust, we formulate a

preliminary definition of trust within the context of a SocialCloud as follows:

Trust is a positive expectation or assumption on futureoutcomes that results from proven contextualised personalinteraction-histories corresponding to conventional relation-ship types and can be leveraged by formal and informal rulesand conventions within a Social Cloud to facilitate as well asinfluence the scope of collaborative exchange.

Trust is a complex construct in which the past (events andexperiences) and the future (plural expectations) are interwo-ven. In regard to future aspects, trust has to be consideredas a multidimensional category, i.e. an attitude of pluralexpectations including competence, integrity and fairness asattributes of the trusted person (see [49] who raised the ideaof multidimensional trust for eBay).

This definition has many implications and demands theprecise explaination of its fundamental aspects (especially thenotions of proven and context) for a better understanding.Human beings always meet in a social context like family,organisationa or communities. People have encounters in in-teraction histories with each other within such contexts. Theexperiences which are made in these contexts function asimplicit or explicit orientations for their future actions andexchanges with others. Above all, institutionalized formal andinformal norms (e.g. reciprocity), which are not universal butinterpreted in different social contexts, affect the constitutionof trust and the formation of possible interactions.

Proven means that ongoing interactions of at least twopeople are based on positive experiences with each other inthe past. As a result, future encounters may become more andmore unquestionable. In contrast, for example if one meets aperson one time and one was disappointed by that person inthis situation, then there might be a level of suspicion on thenext encounter or decision to collaborate with this person.

The contextualisation of encounters and in line with thatof trust as a possible result of positive experiences withinencounters has also to be considered in online contexts whereencounters are taking place not directly (face-to-face) butindirectly, i.e. via technical (social) media. Here encountersand relationships are socio-technical-embedded. Accordingly,in the context of a Social Cloud computing environment, basedon a social network, beside different kinds of technical medi-ated relationships, the inherent algorithms and its componentsare to be set in account for the constitution of trust.

V. SUMMARY AND FUTURE WORK

We have presented trust as a key element for a Social Cloudand argued that we need a thorough understanding of trustand the role it plays. To address this issue, we presentedthe key stages of interaction in a Social Cloud as a frameof reference for trust as an enabler of collaboration. Wethen surveyed trust in collaboration and exchange for onlinecommunities and argued that trust in the context of social-network based collaboration cannot currently be measured,represented or fully understood. Finally, we defined trust for aSocial Cloud and articulated its key aspects. As future work we

will implement a socio-economic/socio-technical trust modelfor a Social Cloud to facilitate empirical studies that canquantify the extent of trust as collaboration enabler.

REFERENCES

[1] Kyle Chard, Simon Caton, Omer Rana, and Kris Bubendorfer. SocialCloud : Cloud Computing in Social Networks. In 2010 IEEE 3rdInternational Conference on Cloud Computing (CLOUD), pages 99–106, 2010.

[2] Kyle Chard, Kris Bubendorfer, Simon Caton, and Omer Rana. SocialCloud Computing: A Vision for Socially Motivated Resource Sharing.IEEE Transactions on Services Computing, 99(PrePrints):1, 2011.

[3] F. Douglis and J. Ousterhout. Transparent process migration: Designalternatives and the sprite implementation. Software Practice & Expe-rience, Vol. 21, No. 8:757–785, 1991.

[4] Matt W. Mutka. Estimating capacity for sharing in a privately ownedworkstation environment. IEEE Transactions on Software Engineering,Vol. 18, No. 4:319–328, 1992.

[5] A. Acharya et al. The utility of exploiting idle workstations for parallelcomputation. In Proceedings of ACM International Conference onMeasurement and Modelling of Computer Systems (SIGMETRICS 97),1997.

[6] C. Haas, S. Caton, and C. Weinhardt. Engineering incentives in socialclouds. In Proceedings of the 11th IEEE/ACM International Symposiumon Cluster, Cloud and Grid Computing (CCGrid), pages 572–575, 2011.

[7] Zhenhua Guo, Raminderjeet Singh, and Marlon Pierce. Building thepolargrid portal using web 2.0 and opensocial. In Proceedings of the5th Grid Computing Environments Workshop (GCE ’09), pages 1–8,New York, NY, USA, 2009. ACM.

[8] Roger Curry, Cameron Kiddle, Nayden Markatchev, Rob Simmonds,Tingxi Tan, Martin Arlitt, and Bruce Walker. Facebook Meets theVirtualized Enterprise. In Proceedings of the 12th International IEEEEnterprise Distributed Object Computing Conference (EDOC ’08),pages 286–292, Washington, DC, USA, 2008. IEEE Computer Society.

[9] Dinh Nguyen Tran, Frank Chiang, and Jinyang Li. Friendstore: Co-operative online backup using trusted nodes. In Proceedings of the1st International Workshop on Social Network Systems (SocialNet’08),Glasgow, Scotland, 2008.

[10] David De Roure, Carole Goble, and Robert Stevens. The design andrealisation of the myexperiment virtual research environment for socialsharing of workflows. Future Generation Computer Systems, 25(5):561–567, 2009.

[11] Gerhard Klimeck, Michael McLennan, Sean P. Brophy, George B.Adams, and Mark S. Lundstrom. nanoHUB.org: Advancing Educationand Research in Nanotechnology. Computing in Science and Engineer-ing, 10:17–23, 2008.

[12] Adam McMahon and Victor Milenkovic. Social volunteer computing.Journal on Systemics, Cybernetics and Informatics (JSCI), Volume9(4):34 – 38, 2011.

[13] Runfang Zhou and Kai Hwang. Powertrust: A robust and scalablereputation system for trusted peer-to-peer computing. IEEE Transactionson Parallel and Distributed Systems, pages 460–473, 2007.

[14] Jordi Sabater and Carles Sierra. Social ReGreT, A reputation modelbased on social relations. SIGecom Exchanges, pages 44–56, 2002.

[15] Sepandar D. Kamvar, Mario T. Schlosser, and Hector Garcia-molina.The Eigen-Trust Algorithm for Reputation Management in P2P Net-works. In Proceedings of the 12th International World Wide WebConference, Budapest, Hungary, 2003.

[16] Li Xiong and Ling Liu. Peertrust: Supporting reputation-based trust forpeer-to-peer electronic communities. IEEE Transactions On Knowledgeand Data Engineering, pages 843–857, 2004.

[17] Mark Granovetter. The Impact of Social Structure on EconomicOutcomes. Journal of Economic Perspectives, 19(1):33–50, 2005.

[18] Kenneth J. Arrow. Gifts and exchanges. Philosophy and Public Affairs,1(4):343–362, 1972.

[19] Robert V. Kozinets. E-Tribalized Marketing? The Strategic Implicationsof Virtual Communities of Consumption. 17(3):252–264, 1999.

[20] M. S. Granovetter. The strength of weak ties. The American Journal ofSociology, 78(6):1360–1380, 1973.

[21] Gayatri Swamynathan, Christo Wilson, Bryce Boe, Kevin Almeroth, andBen Y. Zhao. Do social networks improve e-commerce?: a study onsocial marketplaces. In Proceedings of the first workshop on Onlinesocial networks, WOSN ’08, pages 1–6, New York, NY, USA, 2008.

[22] Joyce Berg, John Dickhaut, and Kevin McCabe. Trust, Reciprocity, andSocial History. Games and Economic Behavior, 10(1):122–142, 1995.

[23] Ernst Fehr and Klaus M. Schmidt. Chapter 8 The Economics of Fairness,Reciprocity and Altruism - Experimental Evidence and New Theories,volume 1 of Handbook on the Economics of Giving, Reciprocity andAltruism, pages 615–691. Elsevier, 2006.

[24] Robert H.J. Mosch. The economic effects of trust. Theory and empiricalevidence. PhD thesis, Tinbergen Institute, 2004.

[25] G.E. Bolton and Axel Ockenfels. ERC: A theory of equity, reciprocity,and competition. The American Economic Review, 90(1):166–193, 2000.

[26] Armin Falk and Urs Fischbacher. A theory of reciprocity. Games andEconomic Behavior, 54(2):293 – 315, 2006.

[27] James C. Cox. How to identify trust and reciprocity. Games andEconomic Behavior, 46(2):260 – 281, 2004.

[28] James S Coleman. Foundations of Social Theory, volume 69. HarvardUniversity Press, 1990.

[29] Niklas Luhmann. Vertrautheit, Zuversicht, Vertrauen. Probleme undAlternativen. In Martin Hartmann and Claus Offe, editors, Vertrauen.Die Grundlage des sozialen Zusammenhalts, pages 143–160. 2001.

[30] A. Giddens. Konsequenzen der Moderne. Suhrkamp-TaschenbuchWissenschaft. Suhrkamp, 1997.

[31] Martin Endress. Vertrauen - soziologische perspektiven. In MatthiasMaring, editor, Vertrauen - zwischen sozialem Kitt und der Senkung vonTransaktionskosten. KIT Scientific Publishing, Karlsruhe, 2010.

[32] M. Endress. Trust and the dialectic of the familiar and the unfamiliarwithin the life-world. In Interaction and Everyday Life. Phenomeno-logical and Ethnomethodological Essay in Honor of George Psathas.Lanham a.o.: Lexington Books, forthcoming.

[33] H. Garfinkel. A conception of, and experiments with, trust’ as acondition of stable concerted actions. In O. J. Harvey, editor, Motivationand Social Interaction, pages 187–238. New York, 1963.

[34] Alfred Schtz and Thomas Luckmann. Strukturen der Lebenswelt. UVKVerlagsgesellschaft, 2003.

[35] Lynne G Zucker. Production of trust: Institutional sources of economicstructure, 1840-1920. Research In Organizational Behavior, 8:53–111,1986.

[36] J. David Lewis and Andrew Weigert. Trust as a social reality. SocialForces, 63(4):pp. 967–985, 1985.

[37] Martin Endress. Vertrauen. Bielefeld: transcript Verl., 2002.[38] Ralf Adelmann. Von der freundschaft in facebook. mediale politiken

sozialer beziehungen in social network sites. In Generation Facebook.ber das Leben im Social Net. Leistert, O. / Roehle, T., Bielefeld, 2011.

[39] J. Derrida and S. Lorenzer. Politik der Freundschaft. SuhrkampTaschenbuch Wissenschaft. Suhrkamp, 2002.

[40] Nicole B. Ellison, Charles Steinfield, and Cliff Lampe. The Benefits ofFacebook Friends: Social Capital and College Students’ Use of OnlineSocial Network Sites. Journal of Computer-Mediated Communication,12(4):1143–1168, July 2007.

[41] Danah Boyd. Friendship. In Mizuko Ito, Sonja Baumer, and Matteo Bit-tanti (Eds.), editors, Hanging Out, Messing Around, Geeking Out: Livingand Learning with New Media, pages 79–115. MIT Press, 2010.

[42] Sherry Turkle. Alone Together: Why We Expect More from Technologyand Less from Each Other. Basic Books, Inc., NY, USA, 2011.

[43] Andreas Wittel. Auf dem weg zu einer netzwerk-sozialitt. In A. Hepp,F. Krotz, S. Moores, and C. Winter, editors, Konnektivitt, Netzwerk undFluss, pages 163–188. VS Verlag fr Soz.wiss., 2006.

[44] Mark Granovetter. Economic action and social structure: The problemof embeddedness. American Journal of Sociology, 91(3):481–510, 1985.

[45] Henning Nuissl. Bausteine des vertrauens - eine begriffsanalyse.Berliner Journal fr Soziologie, 12:87 – 108, 2002.

[46] Christian Stegbauer. Freundschaft als beziehungsform: von der formalensoziologie zu harrison white. In Georg Stanitzek Natalie Binczek, editor,Strong ties / Weak ties. Freundschaftssemantik und Netzwerktheorie.Universittsverlag Winter, 2010.

[47] Lee Rainie, John Horrigan, Barry Wellman, and Jeffrey Boase. Thestrength of internet ties. Available Online at http://www.pewinternet.org,2006.

[48] Bernadette Kneidinger. Facebook und Co. Eine soziologische Analysevon Interaktionsformen in Online Social Networks. Wiesbaden: VS,2010.

[49] Brinkmann and Seifert. ”face to interface”: Zum problem der vertrauen-skonstitution im internet am beispiel von elektronischen auktionen.Zeitschrift fuer Soziologie, 30, 2001.