Abstracts Category : Other

Enhancing trustability in MMOGs environments

by Rui Joo Cardoso

Massively Multiplayer Online Games (MMOGs; e.g., World of Warcraft), virtual worlds(VW; e.g., Second Life), social networks (e.g., Facebook) strongly demand for moreautonomic, security, and trust mechanisms in a way similar to humans do in the reallife world. As known, this is a difficult matter because trusting in humans and organizationsdepends on the perception and experience of each individual, which is difficult toquantify or measure. In fact, these societal environments lack trust mechanisms similarto those involved in humans-to-human interactions. Besides, interactions mediatedby compute devices are constantly evolving, requiring trust mechanisms that keep thepace with the developments and assess risk situations.In VW/MMOGs, it is widely recognized that users develop trust relationships from theirin-world interactions with others. However, these trust relationships end up not beingrepresented in the data structures (or databases) of such virtual worlds, though theysometimes appear associated to reputation and recommendation systems. In addition,as far as we know, the user is not provided with a personal trust tool to sustain his/herdecision making while he/she interacts with other users in the virtual or game world.In order to solve this problem, as well as those mentioned above, we propose herein aformal representation of these personal trust relationships, which are based on avataravatarinteractions. The leading idea is to provide each avatar-impersonated playerwith a personal trust tool that follows a distributed trust model, i.e., the trust data isdistributed over the societal network of a given VW/MMOG.Representing, manipulating, and inferring trust from the user/player point of view certainlyis a grand challenge. When someone meets an unknown individual, the questionis Can I trust him/her or not?. It is clear that this requires the user to have access toa representation of trust about others, but, unless we are using an open source VW/MMOG,it is difficult not to say unfeasible to get access to such data. Even, in an opensource system, a number of users may refuse to pass information about its friends, acquaintances,or others. Putting together its own data and gathered data obtained fromothers, the avatar-impersonated player should be able to come across a trust resultabout its current trustee. For the trust assessment method used in this thesis, we usesubjective logic operators and graph search algorithms to undertake such trust inferenceabout the trustee. The proposed trust inference system has been validated usinga number of OpenSimulator (opensimulator.org) scenarios, which showed an accuracyincrease in evaluating trustability of avatars.Summing up, our proposal aims thus to introduce a trust theory for virtual worlds, itstrust assessment metrics (e.g., subjective logic) and trust discovery methods (e.g.,graph search methods), on an individual basis, rather than based on usual centralizedreputation systems. In particular, and unlike other trust discoveryAdvisors/Committee Members: Gomes, Abel Joo Padro, Freire, Mrio Marques.