Determining Interesting Rules for Many-Valued Concepts using Rule Mining – Information Gain
Abstract
Finding the interesting association rules is very difficult task from enormous collection of raw data. Number of associations among the data can be generated and rules are found using Rule Mining Concepts. Formal Concepts produce number of relationship-concepts which are huge in number but all of them may not be interesting and so finding out of rules that are interesting are required to be found. The main purpose of this paper, to propose a method for discovering novel interesting association- rules in FCA involving multiple-valued concepts using information gain and entropy measure in context rule mining. The present proposed method is implemented and shown by taking real-time examples in day-to-day life.