1 Customer E-loyalty: From an Estimate in Electronic Commerce with an Artificial Neural Fuzzy Interface System (ANFIS) Nader Sohrabi Safa, Maizatul Akmar Ismail Department of Information Science Faculty of Computer Science & Information Technology University of Malaya, 50603 Kuala Lumpur, MALAYSIA Email: sohrabisafa@yahoo.com,maizatul@um.edu.my Tel: +60102402372 and +60133476676 Abstract Companies lose their online customers due to the competitive business environment. Customer loyalty is one of the important topics in the Electronic Commerce (E-commerce) domain. Gaining new loyal customers requires extensive expenditure of time and money. In addition, loyal customers are an important asset for a company, which brings long-term benefits. In this research, a comprehensive conceptual framework is presented that shows E-loyalty based on E-trust and E-satisfaction. The critical factors which influence E-trust and E-satisfaction are classified in organizational, customer and technological groups. Statistical analysis is applied for validity and reliability of the model. Another important method for estimation of uncertain measures is Artificial Neural Fuzzy Network System (ANFNS). E-trust and E-satisfaction data were used as inputs of the ANFIS and the output utilized E-loyalty. The result demonstrated, there is no difference between the aforementioned and the ANFIS model can be used for estimation of E-loyalty in E-commerce. Keyword: E-commerce, E-loyalty, E-trust, E-satisfaction, Artificial Neural Network 1. Introduction Constant business growth is one of the important issues that are guaranteed with loyal customers. Some experts believe that loyal customers are an important asset for every company[1-3].In this research, E- loyalty which is comprised of two important components, E-trust and E-satisfaction, has been presented. Literature review and interviews with experts revealed that customer, organizational and technological factors influence E-trust and E-satisfaction. Data collection tools such as user and expert interviews and questionnaires by means of the Likert scale were used in this regard. This conceptual framework is accepted based on the results of data analysis. In the next part of this research, the application of ANFIS predicted E- loyalty based on two inputs of E-trust and E-satisfaction. Moreover, the usage of the Artificial Neural Network (ANN) predicted E-loyalty based on membership functions [4]. The results show that there is no difference between the results of the ANFIS model and E-loyalty calculated data. 2. Research framework Different experts have paid attention to loyalty in varying aspects. According to Fang, Chiu [5], they discussed E-satisfaction through information, service and system quality. Though, as considered by Chang