Abstract-Cloud systems are in the list of the most recent technologies. It is expected that users move to these new systems rapidly; however, privacy issues make users doubtful to migrate to cloud. They need to give a guarantee that their data and processes are protected enough within the distributed cloud infrastructure; on the other hand, to get better and more special services they have to permit the hosts to access their personal data for personalization purposes. To solve this controversy, in this paper we introduce a new personalization framework and architecture conveys privacy issues. Our proposed architecture mostly involves user side processing and no personal data leaves the client. In this way, the users feel more comfortable to use the cloud system while the quality of service is promoted as well. This method embraces personal data processing agent in the client side through personalization techniques and queries are sent to the hosts in an anonymous format. What is more, our personalization system needs only light processing, able to be run on clients with minimum processing capability, one of the most important issues for cloud users. Another advantage of the proposed architecture is its technology independency that matches the modules with various implementation techniques. We explain our framework and architecture in a "location advising service case study."