Socio-CognitiveGrids方案.PDFVIP

Approximate Profile Utilization for Finding Like Minds and Personalization in Socio-Cognitive Grids Penny Noy and Michael Schroeder Dept. of Computing, City University, London, UK (p.a.noy,msch)@soi.city.ac.uk Summary more satisfying interactions. This scenario is an example of This paper is a tentative discussion of the rationales, roles, finding like minds. methods and drawbacks of employing profiles in the socio- Enhancing the ‘Leaving San Vincenzo’ Scenario In de cognitive grid context, together with the description of a Bruijn and Stathis’ scenario ‘Leaving San Vincenzo’ (de particular method using a lightweight approximation of the Bruijn Stathis 2003), the use of profiles will enable more profile and enabling similarity to be estimated without re- pertinent information to be supplied, because the response vealing the details of the user’s profile. Profile is used in a can be per sonalized to the user. If agent systems are in- general sense to be of a task or a user. The finding of like cluded in the scenario, more possibilities arise. minds relates to the matching of these task- or user-profiles. Profile Data Profiles in Socio-Cognitive Grids The above scenarios need to employ profile data effectively. The development of socio-cognitive grids provides a num- There is an enormous amount of profiling data available, ber of situations in which similarity measurements based but it is currently hardly used. Such data is available to upon profiles could or should play a role: