Analyzing ecology of Internet marketing in small- and medium-sized enterprises (SMEs) with unsupervised-learning algorithm

Internet marketing is a business imperative due to the irrevocable and unstoppable trend of Internet. In this paper, we propose the methodology of unsupervised learning algorithm to apply on the survey data of Internet marketing. Hidden relationships among critical factors in the model are examined using PC-algorithm. To analyze the ecology of Internet marketing in small-and-medium enterprises, similar commercial organizations are grouped into segments by K-means clustering for studying their characteristics. The last methodology is to perform two levels of clustering using self-organizing map and K-means algorithm to save computational cost for massive data instances. The analytical result describes the overview of the Internet usage under stiff-competition business environment which is beneficial to management level to make a more appropriate decision to upload the existing marketing activities online.