Customer-centric data analysis

With the advancement of information technology and declining hardware price, organizations and companies are able to collect large amount of personal data. Individual health records, product preferences and membership information are all converted into digital format. The ability to store and retrieve large amount of electronic records benefits many parties. Useful knowledge often hides in a large pool of raw data. In many customer-centric applications, customers want to find some "best" services according to their needs. However, since different customers may have different preferences to find "best" services, different services are suggested accordingly to different customers. In this thesis, we study models for different customer needs. Besides, customers also want to protect their individual privacy in many applications. In this thesis, we also study how individual privacy can be protected.