Hybrid Clustering System Applied in Patent Quality Management - Take Intelligent Car Industry for Example

Intelligent Car has been the key area of our future technology, In recent years, more and more competitors started to invest one after another. Therefore, it is essential for the researcher to track how to improve their technology and research ability in intelligent car industry area. In this research, the future development of Intelligent Car industry is discussed using patent analytic tools. First of all, related patents are searched and collated by the key words in order to determine twenty key statistic indicators. Then, six are selected among these twenty key statistic indicators by stepwise regression method for SOM (Self-organizing map) clustering use later on. Finally, according to the result from SOM clustering, the patent clusters are defined as high, mid-high, middle, or low value patents. Corresponding forecasting models are specifically designed for each value type of patents in this research for industrial researchers to evaluate strength and weakness of competitors when developing competition strategies.