Evaluating E-Commerce Trust Using Fuzzy Logic

Introduction

Business to consumer (B2C) electronic commerce (EC) has seen a phenomenal growth since the development of the Internet, and there is a growing interest from many organizations to use it as a way to improve their competitiveness and reach a wider customer base. In B2C EC, the concept of trust is crucial because it affects a number of factors essential to online transactions, including security and privacy. It is widely acknowledged that without trust EC cannot reach its full potential (Cheskin Research Group, 1999). Among the most cited concerns of EC customers are the low level of personal data security, inconvenience systems, disappointing purchases, unwillingness to provide personal details, and mistrust of the technology (Cheskin Research Group, 1999; Lewicki & Bunker, 1996; Matthew & Turban, 2001; Mayer et al., 1995; Shapiro et al., 1992).

Kasiran and Meziane (2002) developed a trust model for B2C EC that is based on the kind of information customers are looking for on a vendor’s Website to help them decide whether to engage in a transaction or not. The model identified four major factors that need to be present on a merchant’s Website to increase customers’ trust when shopping online. These factors are: existence, affiliation, policy, and fulfilment. The information the customer needs to collect to satisfy the existence factor include physical existence, such as the merchant’s telephone number, fax number, postal address, mandatory registration, and peoples’ existence. These are known as variables. The affiliation factor looks at third-party endorsement, membership and portal and the policy factor looks at information with regards to customer satisfaction policy, privacy statement, and warranty policy. Finally, the fulfilment factor looks at delivery methods, methods of payment and the community comments. Hence, a total of 12 variables have been identified as summarized in Figure 1.

Figure 1.

The trust model (Kasiran & Meziane, 2002)

Given the large amount of information the model requires, an information extraction system has been developed to automate the data collection process (Meziane & Kasiran, 2003, Meziane & Kasiran, 2005). Indeed it has been reported that users are finding it difficult to identify specific information on Websites (Center for the Digital Future, 2004). In addition, we do recognize that users may not be able to make proper use of the collected information. For this purpose, we developed tools to evaluate the trustworthiness of an EC Website based on the collected information. Two models have been developed in (Meziane & Kasiran, 2005) for evaluating the trust factor; the linear model and the parameterized model. More details about these two models will be provided in the comparison section.

However, for both models, we do recognize that this is not the natural way customers use to evaluate their trust towards online merchants or make the decision to buy or not. As with any other business transaction, customers develop in their mind some sort of ambiguity and uncertainties when purchasing online (Mohanty & Bhasker, 2005). The customer may wish to classify the merchants using different preferences or take into accounts other parameters such as the cost or the brand of the product. The decision to buy or not to buy online is often based on user’s human intuitions, common sense, and experience rather than on the availability of clear, concise, and accurate data (Akhter, Hobbs, & Maamar, 2005). In this article, we develop a new trust evaluation model using fuzzy reasoning to evaluate the trust factor as it allows the encoding of the information available on the merchant’s Website in a form that can be used to reflect the way customers reach the decision to engage in an EC transaction.