Author privacy, data fabrication, and Knowledge Discovery in Databases

The problem of data fabrication, due to heightened consumer concerns about privacy, is on the rise. The unique characteristic of the Internet, anonymity, is a probable contributor to the intention of users to fabricate information. We propose a technological solution to this problem based on the deployment of Knowledge Discovery in Database (KDD) systems to learn discrimination functions that discriminate between correct and fabricated data. These discrimination functions can then be used to form filters that remove falsified data from marketing data. That such discrimination functions are possible is due to the characteristic form falsified data takes. The greatest hurdle to implementing this approach is the availability of data labeled as "falsified" and "correct." However, the proposed technological solution offers potential to marketers and businesses alike © 2006 IEEE.