Kansei Engineering Research on Character Orientation of Virtual Assistant
| Autor: | CHOU, PEI-YAO, 周佩瑤 |
|---|---|
| Rok vydání: | 2019 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 107 Along with the rapid growth of artificial intelligence and voice-interface technology, more and more studies have been conducted to explore the psychological impact of machine emotions on its users. In the past, Kansei Engineering has been used mostly in product design to help designers master the use of design elements through quantified and specific practices. With a focus on virtual assistant products with human-computer communication features, this study explores users’ images of virtual assistant characters based on Kansei Engineering, produces an semantic differential scale suitable for evaluating virtual assistant characters, and summarizes the key elements in designing virtual assistant characters. First, this study collected emotional vocabularies, targeted the virtual assistant products with the highest market shares, and selected 15 groups of opposing adjectives from the vocabulary collected from the factor analysis of interviews with three experts. The study then invited 21 volunteers to fill in the semantic differential scale. Lastly, design elements with relatively higher points were screened out by applying Quantification Theory I to these results. The results of this study suggest that a “voice plus light” is well-suited for signaling the “Wake-up Response” of virtual assistants, as users respond to this with comments like “amiable, polite, and attractive.” Using a female voice will make the virtual assistant appear “smart, natural, and welcoming.” As for the “Visual Appearance” of virtual assistants, assistants with a character image appear more “scientific and stereoscopic” than those without a character image. For the “social conversation” feature of the virtual assistants, those with both a “task-oriented approach” and “emotional appeals” were most welcomed by users, receiving comments like “helpful, professional, and friendly.” In the “personality” section of the assessment, “round characters” tend to give users the impression of being “compliant and considerate,” “vivacious and talkative,” and “humorous.” Based on the abovementioned research results, this study verifies the feasibility of utilizing Kansei Engineering in research on voice user-interface technology. It also summarizes users’ inner feelings about different virtual characters, providing a reference for future designs of virtual assistant products. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|