聊天機器人性別提示對使用者在網路購物中的信任度與個人感受影響

The Impact of Chatbot's Gender Cues on User's Trust and Personal Feeling in Online Shopping

陳繹介(Yi-Chieh Chen)

指導教授 ： 游曉貞

國立臺中科技大學/設計學院/多媒體設計系碩士班/碩士(2018年)

摘要

近來伴隨著科技的發展，進而帶動著人工智慧的進步與興起，各行各業陸續嘗試將人工智慧技術帶入生產行為當中，而其中又以服務業最為普遍，人機對話成為人機互動中的一項最新趨勢，語音助理、聊天機器人皆為其代表性產品，許多諸如Siri 與Messsnger bot等虛擬助手紛紛發布問世，如何透過聊天機器人在電商服務中取得客戶的信任度成為一項新的課題。過去人們普遍傾向將人工智慧助理的社會性別認定為女性，而在性別化創新這一概念提出後，諸多人工智慧產品亦紛紛加入了性別屬性的考量，然而，不同於語音助理，訊息式聊天機器人在互動上，並不具備如語音一般強烈的性別提示，僅單純以文本對話的方式與使用者進行互動，因此，在透過聊天機器人的文本對話中提供性別提示的方式，在網路購物情境中探討使用者對於不同性別屬性的聊天機器人的信任度與個人感受程度。本研究中實驗分為兩部分，首先透過前導實驗透過134份問卷蒐集的方式確立頭像性別後，將聊天機器人的性別提示作為自變項，參與者分別與不同性別提示組合的聊天機器人進行互動。正式研究中邀請男、女各60名，共120位實驗參與者，將平均分配於四種實驗條件中：(A)男性姓名￫男性頭像、(B)中性姓名￫男性頭像、(C)中性姓名￫女性頭像、(D)女性姓名￫女性頭像等四種實驗條件，並以模擬使用聊天機器人進行網路購物之情境，透過問卷填寫的方式，了解參與者對於聊天機器人的信任度與個人感受，以探討不同性別提示下，參與者對於聊天機器人的信任度與感受是否具有差異。研究結果表示，聊天機器人性別與參與者性別差異對於信任問卷中的正直構面與其餘構面部分問項中都具有顯著性差異，同時在年齡層分布中，亦有發現對於聊天機器人信任與感受有顯著性差異存在。結論中根據實驗分析結果，對於未來聊天機器人之設計開發時有關性別提示設定進而提出建議。

關鍵字

對話式使用者介面 ； 聊天機器人 ； 性別提示 ； 信任度 ； 個人感受

並列摘要

In recent years, due to the rapid development of science and technology, and thus the advancement and rise of artificial intelligence, different companies have gradually introduced artificial intelligence technology to their products and services. Among them, AI is used most frequently in the service industry, and the human-machine dialogue has become the latest trend in human-computer interaction. Voice assistants and chatbots are among the representative products. Many virtual assistants such as Siri and Messenger bots have come out one after another. Therefore, how to gain customer trust in e-commerce services through chat bots has become a new topic. In the past, people generally preferred to identify the gender of artificial intelligence assistants as female. Many artificial intelligence products also had considered to AI’s gender issue after the concept of Gendered Innovations was proposed. However, unlike voice assistants,,message-based chatbot does not have the obvious gender cue such as like voice. It only interacts with user with text message and Emoticons. Therefore, the purpose of this paper is investigating on trust and feeling about different gender chatbot in on-line shopping. Therefore, the purpose of this paper was investigating on trust and feeling about different gender chatbot in on-line shopping. At first,we invited 134 participants for the pre-test to define the gender of the chatbot, and In the formal experiment, then, we invited other 120 participants who were assigned to the four experimental conditions: : (A) male avatar with male name , (B) male avatar with neutral name, (C) female avatar with neutral name, (d) female avatar with female name" → female avatar and inteacted with them.After participant wanted to finish, we asked them evaluated the trust and feeling with the scale. The results of the study indicate that the gender difference between the gender of the chat robot and the gender of the participants is significantly different between the positive facet and the other facet in the trust questionnaire. At the same time, in the age distribution, the trust and trust of the chat robot are also found. There are significant differences. According to the experimental analysis results, the conclusions about the gender prompt setting in the design and development of the future chat robot are made.

並列關鍵字

Conversational User Interface ； Chatbot ； Gender Cues ； Trust

參考文獻

英文文獻

1. Abu Shawer.B,& Atwell.E.(2007).Fostering language autonomy through adaptive conversation tutors. Proceedings of 2007 Corpus Linguistics Conference.

2. Armstrong, C. L., & Nelson, M. R. (2005). How newspaper sources trigger gender stereotypes. Journalism & Mass Communication Quarterly, 82(4), 820–837.

3. Armstrong, C. L.& McAdams, M.J.(2009,July) Blogs of Information: How Gender Cues and Individual Motivations Influence Perceptions of Credibility.Journal of Computer-Mediated Communication,Volume 14, Issue 3,Pages 435–456.

4. Adams, B. D., Bruyn, L.E., Houde, S., & Angelopoulos, P. (2003). Trust in Automated Systems Literature Review (Report No. CR-2003-096). Toronto, Ontario, Canada: Department of National Defense.