使用N組連結平均法的階層式自動分群

A Novel Algorithm Using N-link Average for Hierarchical Automatic Clustering

行動電信技術的發展，促使行動廣告成為重要的個人化廣告媒介。本研究探究個人化行動廣告設計要素的相對重要性、與適用的屬性產品。本研究回顧文獻與實務案例，確認個人化行動廣告設計要素的完整性。延續模糊德菲爾法（Fuzzy Delphi Method）所篩選出的重要行動廣告設計要素，進行模糊多準則決策（Fuzzy Multi-criteria decision making, Fuzzy MCDM），以確認要素的相對重要性，並評選出適用的產品屬性。本研究發現實體商品比服務更適合使用個人化行動廣告，其中，價格與促銷模式為最重要的設計要素。本研究結果可減少行動廣告相關業者之間的認知差異，並增加合作時之共識與合作之有效性，以增加資源使用之效率。

The maturity of mobile broadband technology promotes the mobile advertising as an important personalized adverting media. This study explored the relative importance of the key factors of the personalized mobile advertisement and the suitability for product characteristics. Through literature review and practical case studies, this research identified important factors in designing a personalized mobile advertising message. This research employed the key personalized mobile advertising factors identified by fuzzy Delphi method to evaluate the relative importance of the factors and product characteristics. By using MCDM (Multi-Criteria Decision Making) research method, our research found that price and promoting are the most important factors for designing personalized mobile advertisement. Comparing with service, physical product is better suited for mobile advertising approach. This research aims to help players in the mobile advertisement value chain to enhance consensus in establishing a collaboration model.