在电子商务盛行的时代，由于规避了建立实体客户服务中心和顾客的交通成本，基于互联网的服务形式已成为传统的实体客服中心的一个替代形式。然而，在网络上由于缺少客服人员与用户间的面对面交流，降低了沟通的有效性以至于影响了服务质量。因此，为了同时实现网络服务的低成本运营和高质量的服务，有必要开发新的服务渠道。三维互联网空间（虚拟世界）便可能成为客户服务的一个新媒介：用户通过其网络虚拟替身，可以实现基于虚拟地理位置的、实时的"面对面沟通"。本项目是以三维虚拟世界作为潜在服务平台的探索性研究：以媒体丰富度理论和社会存在感理论为基础建立的关于服务质量的概念模型，测量不同服务场景下用户评估的服务质量，1.比较在三维虚拟世界和普通网络服务的质量和客户满意度；2.剖析在不同服务渠道中，决定客户服务质量的哪些关键维度显著影响客户满意度；3.探讨三维虚拟世界作为新型服务平台的可行性，及其在不同领域提供服务的优劣势。

In the Internet era, web-based services have become a convenient alternative to physical customer service interactions, as it saves the construction and operation cost of the customer service center and customers' transportation cost purchasing the products or service. However, according to media richness theory, lack of face-to-face interaction makes web service communication inefficient. Thus a new channel of online service is called for to satisfy customers with both convenience and efficiency. The 3D virtual worlds become a potential platform that offers customer service, where users can communicate "face to face" via their representative avatars. This paper is an exploratory study of proposing a potential channel of customer service in 3D virtual worlds. Theories of computer display technology, communication, and psychology, such as social presence theory, media richness theory, and social identity model of deindividuation effects (SIDE model), are applied to address how a 3D virtual world impacts users' sense of presence, and their perception of customer service quality. We propose a conceptual model of service quality to compare the quality of customer service and users' satisfaction in 3D virtual worlds to that of web-based services, to identify factors incluencing service quality and user satisfaction in 3D virtual worlds, and to investigate the advantages and disadvantages of customer service in both platforms with different service tasks. We design an experiment in one of the most famous virtual worlds Second Life and set up a mock-up website to collect data in a post-study questionnaire to test our conceptual model. Managerial applications will be discussed based on the results of data anaysis in this project.