近年來，「社群媒體」與「共享經濟」結合的行動電商發展方興未艾， 其中以UBER 進軍各國計程車市場所引起的風潮為較知名案例。目前UBER 在臺灣營運是否合法的爭議不斷，尤其涉及叫車方式、彈性費率、司機乘 客互評、稅制等課題，皆面臨適法性之挑戰，因此，若能經由社群媒體挖 掘對於UBER 相關課題之文本進行情感分析，進而了解民眾對UBER 適法 性的情感態度，則可提供政府推動多元化計程車服務之參考。本研究使用 爬蟲系統蒐集UBER 網路文本進行挖掘，並建構三元決策情感分析模式， 該模式係將情感傾向區分為正面、中立與負面情感。實證結果顯示，討論 聲量前三名的營運話題分別為：營運制度、取締與抗議以及稅務與收費。 民眾對於UBER 在臺營運之網路平台服務的評價，整體情感趨勢係以正面 偏中立為主，高於負面情感傾向。整體負面情感趨勢則為UBER 營運合法性之爭議，討論量以非法營業與未盡繳稅義務最多。因此，經由社群媒體 挖掘而得出之共同看法是UBER 應先取得合法執業許可及繳納相關稅費， 方可在臺營運。

Mobile commerce business combining sharing economy and social media are now in the ascendant, especially UBER launches into taxi markets worldwide which becomes the most attractive case study in research literatures. Nowadays, issues of legal operations concerning service calling, flexible fares, rating system between drivers and users and tax are still in dispute for UBER in Taiwan. It will be very helpful for Taiwan’s authorities and taxi operators if users can understand and accept similar services like UBER by using social media mining and sentiment analysis. Crawler systems were used to collect all possible text data from social media. Opinion mining and sentiment analysis were sequentially performed in this study. A model based on the theory of three-way decisions was used for sentiment analysis which included three sentiment zones: positive, negative and neutral. Results of the empirical study showed that the three topics about UBER operations concerning operation mechanism, regulation and protest and taxation problems were discussed extensively in Taiwan. The hottest topic was mobile platform services offered by UBER which won positive and neutral orientation tendency more than negative one significantly. The whole negative orientation tendency focused on dispute of UBER’s illegal operations in Taiwan. It is evident that the first priority for UBER is to apply legal permission and pay tax arrears as soon as possible if UBER attempts to continue operations in Taiwan.