Minimum explosure in application form relying on a classification processes

Application forms are often used to collect personal information, and subsequently to tailor these services to the applicant's specific situation. The data collected must be reduced to a minimum size in order to serve the objective process. However, no technique exists today to minimize information used by complex decision processes based on classifiers. In this article, we extend our previous results by presenting the generalized Minimum Exposure problem. We show that this problem is NP-hard and discuss its resolution. Experiments show a big reduction in data transmitted by using our technique.