隨網際網路應用技術發展越趨成熟,「網上社群群聚」已經不是單一現象,網路社會是一個新的、更具(更容易找到)共同性與關聯性的社群社會,據此,業界從事行銷推廣的管道終將伸往這些網路族群。 國內研究討論利用網路工具或網上社群運作的行銷議題時,多討論單一工具的價值,而針對整體網上社群行為的討論與探索則寥若晨星,故本研究欲以「可形成網上社群的工具」為研究根本,討論使用者網上社群行為及購物的資訊參考行為,以替網上社群行銷增加可考慮的議題與構面。 本研究利用SPSS Clementine軟體,以「資訊互動的積極程度」與「社群互動的積極程度」為構面,以二階法將網上社群使用者分為「交誼型」、「資訊型」、「工具型」與「八卦型」四種,以資料採礦(Data Mining)技術討論四群使用者對網上社群工具的使用價值、社群感與資訊影響力的關聯知識,並據此對廠商加入網上社群行銷的操作手法進行提案。 Since the development of internet applications matured, social network is a new social mode that is easier for searching the cyber community, and it is forward to create an innovative source for advertising and marketing to the business. Mostly, researches in Taiwan in relation to the behavior or marketing on online community tools are focus on either advertising or marketing, without a whole picture. Thus, this study investigates the on-line community tools as a research subject, and explores the users’ online and purchase behavior in the cyber community. In addition, this study implements the SPSS Clementine as a data mining approach, to categorize four kinds of online users’ behavior by using a two-step method. Finally, this article suggests that online users those who join communities of commerce and their purchase behavior are critical knowledge for considering possible business models and proposals. By doing so, in terms of profit model, knowledge extraction from subjects of four clusters might be a contribution to this research issue.