期刊名称：International Journal of Asia Digital Art and Design Association
电子版ISSN：1738-8074
出版年度：2013
卷号：17
期号：2
页码：38-43
DOI：10.20668/adada.17.2_38
语种：English
出版社：アジアデジタルアートアンドデザイン学会
摘要：This research intends to design better agent interfaces for users to enable superior interaction during language practice.Subsequent to previous two surveys [1][2] that investigated the affective feelings of users towards different agentinterfaces, the outcomes that resulted from the two surveys have been adopted in conjunction with some existingguidelines in order to develop this multi-agent website. The development of this multi-agent website for inexperiencedEnglish learners has been developed with respect to the research results of the previous two surveys [1][2]. These resultsinclude the preferred agent interface design for language learners and the facial expressions of agents that attract users toengage in the learning tasks. Therefore, this is a website designed with preferred agent interfaces that is likely to attractusers to further their engagement in the language practice tasks. The website also incorporates ten characters equippedwith six basic expressions as well as one neutral face, and these were created as CompanionBots.Moreover, this website provides some significant design, computer science and psychological challenges in terms of agentcharacter design (user impression in a learning task), agent emotional expression design (evoking users‟ affectiveresponses), the Emotional Keyword Filter (emotional keywords lead to agent expression exchanges), device compatibility(suitable use on computers and smart mobile devices) and the IELTS (International English Language Testing System)-Bot(designed for IELTS rehearsal use). In truth, this website has been devices via certain computer science techniques andpsychological theories. Also, this website intends to implement these design guidelines into practical use in order todevelop better agent interfaces.
关键词：Affective Agent Interface;Embodied Conversational Agents (ECAs);Facial Expressions of Agent
