| 摘要: | 客戶服務是企業與消費者的最直接的溝通橋樑，他們在顧客眼中代表整個企業與組織，好的客戶服務能夠建立良好形象與信譽，並有助於提升顧客滿意度。而近年來隨著人工智慧的發展，為了提供顧客最即時且有效的回應越來越多企業導入AI客服，由於AI客服所能處理的問題有限，有時必須透過真人客服才能加以解決，因此真人客服與AI客服共同完成客戶服務的人機合作更是一個重要的管理與研究課題。本文以創新擴散理論為基礎建構擴充版的科技接受模型，並以大學學生為對象，除了分析影響消費者使用AI客服的相關因素外，並藉由因素分析技術來探討消費者在不同購物情境下使用AI客服時，哪些因素將扮演關鍵核心角色。 本研究問卷發放期間為2018年12月12日至12月19日，有效問卷總共回收411份。研究結果發現消費者使用AI客服的原因主要為能節省等待時間與有效解決問題，建議業者可以透過自然語言理解、深度學習和情感識別技術的AI雙向對話式交互系統，能真正理解用戶表達語義，利用反饋數據來提高AI客服回答的準確性和服務效果。而透過人機合作的方式與提供多渠道溝通的整合性平台，將進一步可以解決目前AI客服無法有效回應消費者的問題。另外，業者必須針對各種不同屬性的創新採用者給予不同的協助，如利用顧客的瀏覽記錄、行為習慣等資料進行深度學習，能夠精準的建立用戶畫像圖譜，以提供更個性化需求的服務。
Customer service is the most direct bridge of communication between enterprises and consumers. They represent the whole enterprise and organization in the eyes of the customers. A good customer service can establish a strong corporate image and reputation, and is conducive to increased customer satisfaction. In recent years, with the development of artificial intelligence, in order to provide customers with the most immediate and effective response, more and more enterprises have introduced AI customer service. However, AI customer service can only handle a limited amount of customer problems, and sometimes the problem can only be resolved through a live customer service agent. Thus the human-machine cooperation between live customer service agents and AI customer service to resolve customer problems is an important management and research topic. Based on the innovation diffusion theory, this paper constructs an extended version of the technology acceptance model, and is targeted at university students. In addition to analyzing the factors which affect consumers' use of AI customer service, factor analysis technique is also used to explore and understand the factors that play a key role in consumer behaviors under different shopping situations. Research has found that consumers use AI customer service mainly because they can save waiting time and solve problems effectively. Therefore, it is suggested that the dealers can use the two-way interactive conversation system, which covers natural language understanding, in-depth learning and emotion recognition technology to truly understand users' communicated semantics. Furthermore, they should use feedback data to improve the accuracy and service effectiveness of AI customer service responses. Through the human-machine cooperation approach and the integrated platform that provides a multi-channel communication, this will further solve the problem of AI customer service not able to respond effectively to consumers. In addition, the dealer must provide different assistances according to the different attributes of these innovative adopters, for example, using the customer's browsing history, behavioral habits and other required materials in order to conduct an in depth learning. This way they can accurately establish user portrait graph spectrum so as to provide a more personalized services. |
