Abstract
This article briefly describes natural induction approach to knowledge discovery, and then applies it to the problem of bad habit relapse prevention by analyzing patients' diaries. Natural induction seeks patterns in data that are in forms easy to understand and interpret, because they resemble those in which humans represent knowledge, such as natural language descriptions and visual forms. The application of natural induction to the problem of bad habit relapse has produced patterns easy to understand, in some cases of surprising simplicity.