- 學位論文
智慧型手機操弄評論辨識之研究
A Study on Identifying Review Manipulation of Smartphones
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
摘要
隨著Web2.0發展，網路的強大影響力改變了整個社會的溝通模式。使用者利用網路平台所提供的討論機制，例如U-Car、Facebook、Mobile 01、PTT…等平台，讓大家可以提供自身意見與訊息交流。現今許多消費者在購買商品之前會根據網路上的部落格、社群網站、電子商務網站等平台，尋找他們想要購買的商品資訊並做為購買的參考依據。因此，在這樣的新興電子商務環境底下，線上消費者的評論一直扮演著重要的角色。然而，有部分公司企業試圖利用操弄評論來提升銷售額，但消費者面對著因為網路便利所帶來的大量訊息，很難辨別評論其真偽，只能依靠自身經驗來辨識。因此，本研究依文獻探討結果，定義了包含情感語意、產品特徵、專有名詞等15項潛在操弄特徵屬性。並導入這些評論操弄資訊，支持向量機 (Support Vector Machine, SVM)與利用決策樹 (Decision Tree, DT)，試圖提升操弄評論分類績效。另外，本研究使用相關分析(Correlation analysis)、決策樹萃取之規則與植基於倒傳遞類神經網路 (Backpropagation network, BPN)特徵選取法，以辨識本實驗所提出評論操弄特徵之重要度。最後以手機產品操弄評論實例，驗證本研究所提方法之有效性。
並列摘要
Following the development of Web 2.0, the powerful influence of the Internet has changed the communication mechanisms of our society. Internet users used the online communication platforms such as U-car, Facebook, Mobile 01, PTT…etc. to share their opinions and exchange their own experiences. Now, there are lots of consumers depend on the information posted in blogs, social networks, e-business websites and so on to find the information they need before they make the purchase decisions. Therefore, the online consumers’ reviews play an important role in this new e-commerce environment. Consequently, some enterprises attempted to increase the sale volume by manipulating online users’ comments. But, it’s difficult to be identified by consumers. They only can recognize the comments by their own experiences. Therefore, after viewing available literatures, this study defined fifteen features including sentiment, product features, expertise, and so on. By introducing the review manipulation information, we try to improve the classification perforamcnes of Decision Trees (DT) and Support Vector Machines (SVM). In addition, this work used correlation analysis, the extracted rules by decision trees, and neural networks based feature selection method to identify the importance of the defined 15 potential review manipulation attributes. Finally, two real cases of manuplated reviews about smart phones to evaluate the effectiveness of our methods.
參考文獻
[2] 王堅斌 (2003)，使用分類結合法與決策樹判別忠誠顧客，碩士論文，國立成功大學資訊管理學系，台南。
[8] 洪士程、林淑芬 (2013), 應用資料探勘技術於國小教師工作壓力之研究~以彰化縣為例，2013資訊科技國際研討會，霧峰，台中。
[16] 孫昆正 (2012)，發展不平衡語意分類之研究，碩士論文，朝陽科技大學資訊管理所，台中。
[2] Anderson E.W. (1998), “Customer Satisfaction and Word of Mouth,” Journal of Service Research, vol. 1, no. 3, pp. 5-17.
[3] Arndt J. (1967), “Role of product-related conversations in the diffusion of a new product,” Journal of Marketing Research, vol. 4, no. 3, pp. 291-295.