Interesting Knowledge Exploring for Product Component Selection

During the product configuration design stage, it is important to find useful knowledge to reduce the dedicated efforts of engineers. To fulfill this task, the quality and interestingness of knowledge plays the decisive part. This paper utilizes Apriori algorithm to mine rules from historical data base. In addition the traditional criterions support and confidence, the criterion interestingness is also applied to filter rules that might misguide design engineers. The validity of the proposed model is illustrated by a case of electrical bicycles.