ILP for cosmetic product selection

In this chapter, we design a real-world machine learning system using a smartphone. This system can acquire images taken with the camera of a smartphone using learners (ILP and SVM) and automatically diagnose the new image. To develop this system, we implement an image analysis function and a classifier function using learning rules (learned ILP) and learning data (learned SVM) in the smartphone. With this system, the user can collect image data with a smartphone camera and diagnose the new image data according to the learning rules and data in the smartphone. In this chapter, we apply this system to a cosmetics recommendation service and demonstrate its effectiveness by improving the user service.