摘要：It is often d esirable to b e able to gu arantee the in teg rity of h istorical data, en suring that any su bsequen tm odificatio ns to the data can be d etected . It wou ld b e esp ecially con venient to extend s uch p roo fs ofin teg rity to certain comp utations p erform ed later usin g the histo ric data. Ho wever, curren t Recom menderSy stem s Architectu re is not su itab le for u se with extensiv e and s en sitive u ser profile d ata. Thu s, wep ro pose an ap proach to agent-based Information Filterin g resultin g in an architectu re pres erving the p riv acyo f all participants. The pro pos ed solution co vers trust relation ship s b etween particip an ts an d utilizesp riv acy-preservin g imp lem en tatio ns o f ex isting filterin g techniq ues
