社群聊天機器人互動率探究與使用者行為分析

聊天機器人的概念發展至今已有七十年的歷程，隨著使用者的使用習慣轉變及行動裝置蓬勃發展，結合社群媒體之社群聊天機器人也日漸活絡，發展出各式各樣的應用情境，不但使對話式商務興起，更讓使用者與聊天機器人的距離越來越近，而社群聊天機器人即時且容易操作的特性，也被運用於社群媒體之客服服務或娛樂及遊戲服務上。在現今與社群媒體密不可分的世代，互動率及互動體驗更是成為一大評估社群媒體成效的重要指標，也因此開始有經營者將社群聊天機器人導入社群媒體中，藉由社群聊天機器人的力量提升社群媒體之互動率。在各式各樣的應用中，娛樂及遊戲型社群聊天機器人已有提升社群媒體互動率之實，但卻缺乏相關研究文獻以了解背後之歸因，而在進行服務流程及使用體驗優化時也缺乏相關依據。因此，本研究在進行互動率探究之外，也納入其他互動相關概念，與三位社群專家進行半結構式訪談，並搜集、觀察、整理實際一社群聊天機器人相關之數據指標為基礎，針對娛樂及遊戲型社群聊天機器人進行互動率之探討；也利用問卷調查法搜集346份問卷，以人機互動量表檢視使用者的互動感知程度和與再互動意願之間的關係；並實際製作社群聊天機器人貼文，與7位受測者進行測試、訪談與分析，以科技接受模型理論為基礎，探究使用者對於再互動之行為意向；除此之外，更透過與使用者的對話，實際繪製娛樂及遊戲型社群聊天機器人之使用者旅程地圖，以此作為服務及使用體驗優化之基礎。研究結果發現，娛樂及遊戲型社群聊天機器人可為品牌及其粉絲專頁帶來正向影響；「娛樂感」與「感知挑戰」為使用者決定是否進行再互動之考慮因素；另外，若能在滿足認知易用性及認知有趣性後，再額外滿足認知有用性，將能夠在使用者心中留下深刻印象，發揮價值作為使用者的社交資本；而在使用者類型分眾上，可分為連結共鳴型、自我滿足型、理性評估型、社交目的型等四種類型；在服務流程上，最需要改進的部分在留言回覆、同意GDPR、再互動意願、下次推播再互動等階段，應思考如何降低使用者的未知焦慮。

關鍵字

社群聊天機器人；互動率；使用者行為；使用者體驗；科技接受模型

並列摘要

The concept of chatbot can be traced back to 70 years ago. With user habits changing and mobile devices establishing, chatbot designed for social media applications is now flourishing, with multiple functions being developed for various scenarios. Chatbot commerce is a rising business as users’ willingness to engage with chatbot increases. Chatbot’s accessibility and immediacy have been widely utilized in customer service for social media and entertainment purposes. As social media becomes a crucial part of our day-to-day life, engagement and interactivity are deemed vital when evaluating social media performances. Hence, chatbot is now utilized in social media operation as a means of optimizing interactivity with users. Among the various applications, chatbot for entertainment and gaming purposes have shown results in optimizing user interactivity. However, there is a lack of related research to analyze the causes behind such outcomes. References are also hard to acquire when utilizing chatbot or optimizing the user experience. The purpose of this study was to research the engagement of chatbot services. Other interactive concepts were also utilized in the study as well. Semi-structured interviews were conducted with three social media operators to gather and monitor the data, specifically focusing the engagement rate from chatbots used for entertainment and gaming purposes. 346 questionnaires were also collected, detailing surveyors’ human-computer interaction experience and their willingness to repeat the interactions. Chatbot experience were recreated for seven testees for interview and analytical purposes, using Technology Acceptance Model to analyze users’ willingness to repeat the chatbot experience. Furthermore, user journey maps were developed to exemplify their experiences with chatbots for further optimization purposes. The study indicated that chatbots for entertainment and gaming purposes can bring positive influence for brands and fan pages, while “entertainment” and “perceived task challenge” are considered determining factors for users to repeat the experience. In addition to sustaining users ’perceived ease of use and perceived enjoyment, perceived usefulness can enhance the experience and further the service as a social capital for the users. Four categories can be sorted based on the users’ demographic: relatability, satisfaction, rationality and sociability. In service produces, findings suggested that the parts that require the most improvement are the reply messages, GDPR agreement, willingness for repeated interaction and willingness to replicate actions when receiving further notification. The study also indicated that minimizing users’ anxiety when engaging with unknown service is crucial for optimizing chatbot experience.