RHadoop-based fuzzy data mining: Architecture, design and system implementation

Data mining is a challenge for end-users, which requires knowledge and skills on business domains, data mining algorithms and software development. In response to the challenge, we have proposed, designed and implemented a novel data mining system named RFDM (RHadoop-based Fuzzy Data Mining), which supports fuzzy data mining process and experience with user convenience and reduced cost. The system is capable of supporting fully-automated data mining life-cycle activities, with limited user interactions in dataset uploading and data mining configuration. In addition, a RHadoop-based framework has been integrated, which meets the requirements of large-scale datasets in data mining. Experiments have indicated that the RFDM system achieves enhanced performance while supporting fuzzy data mining.