- 學位論文
導航語音助理個性對使用者好感度與滿意度之影響
The Impacts of Voice Navigation Assistant Personality on User Favorability and Satisfaction
摘要
科技日異月新，人工智慧（Artificial Intelligence, AI）技術逐漸成熟，語音助理產品逐漸出現在人們生活當中。啟動裝置中的語音助理，使用者即可透過語音互動來取得資訊或完成例如：查詢天氣狀況、播放音樂、設定行事曆…等任務。不少研究發現塑造語音助理的個性可以給予使用者更好的體驗，因此語音助理的個性對於語音互動設計成為了重要的課題。在語音導航的部分國內也比較少有人去研究，不管是語音助理的形象樣貌賦予語音助理特徵，或者是賦予個性化的聲音都鮮少有一個設計的準則。也較少有針對不同的情境下選擇語音助理個性方面的知識。因此本研究透過兩階段的實驗設計，探討車用語音導航助理個性對於使用者好感度與滿意度的影響。研究問題有三點：(一) 使用者是否能透過聲音來感知不同語音導航助理的個性差異？(二) 在語音導航情境下，不同語音助理個性對於使用者好感度的影響？(三) 在語音導航情境下，不同語音助理個性對於使用的滿意度的影響？第一階段實驗探討如何利用聲音元素的組合塑造語音助理的個性。自變項為性別元素(男生、女生)、語速(快、慢)、音調(高、低)，依此設計2x2x2的組內設計實驗，並透過Youtube平台播放。依變項則是依據五大人格模型理論的構面，以五點量表方式透過線上問卷平台讓受測者進行填寫。第二階段的正式實驗則根據 第一階段結果，挑選出兩個個性差異較大的個性聲音組合 (內向、外向)，作為語音助理的個性提示。本實驗共招募72位受測者，每位受測者都單獨進入實驗場域和語音導航助理進行互動體驗。實驗任務完成後讓受測者填寫對本實驗語音導航助理的好感度與滿意度量表。研究結果發現：(一) 透過不同聲音元素的組合是可以影響使用者對語音助理個性的感知；(二) 在語音導航情境下，使用者對不同個性的（內向、外向）語音助理的好感度並沒有顯著影響；(三) 在語音導航情境下，使用者對不同個性的（內向、外向）語音助理的滿意度並沒有顯著影響。因此，內外向個性針對語音導航助理的使用情境不會有好感度與滿意度影響對此本研究也給出相對應的建議與研究結論。
並列摘要
With technology advancing rapidly, artificial intelligence (AI) has become more mature, and voice assistants are gradually becoming a part of people's lives. Many studies have found that shaping the personality of voice assistants can provide users with a better experience, making the personality of voice assistants a critical aspect of voice interaction design. However, there are few established guidelines for shaping the personality of voice assistants. Therefore, this study conducted a two-stage experimental design to explore the impact of in-car voice navigation assistant personalities on user likability and satisfaction. The research questions are as follows: (1) Can users perceive differences in personality among different voice navigation assistants through their voices? (2) How does the personality of voice navigation assistants affect user likability in the context? (3) How does the personality of voice navigation assistants affect user satisfaction in the voice navigation context? The first stage of the experiment investigated how to shape the personality of voice assistants using combinations of voice elements. The independent variables were gender elements (male, female), speech speed (fast, slow), and pitch (high, low). The dependent variables were based on the five-factor model theory, and participants were asked to fill out a five-point scale questionnaire through an online survey platform. The second stage of the formal experiment selected two voices (introverted and extroverted) with more significant personality differences based on the first stage's results as the navigation assistant's personality cues. A total of 72 participants were recruited for this experiment, and each participant interacted individually with one of the voice navigation assistants in a simulated driving setting inside the laboratory. After completing the tasks, the participants filled out the likability and satisfaction scales for the voice navigation assistant. The findings showed that: (1) Voice elements can influence users' perceptions of the personality of the voice assistant; (2) In the context of voice navigation, users' likability of different personality (introverted and extroverted) voice assistants did not significantly differ; (3) In the context of voice navigation, users' satisfaction with different personality (introverted and extroverted) voice assistants did not significantly differ. Therefore, contrary to many personality literatures, the personality of voice navigation assistants, whether introverted or extroverted, does not significantly affect likability and satisfaction in the usage context of this study. This study also gives corresponding suggestions for future research