Assessing trust in e-Commerce website based on ranking of trust attributes

Abstract

Trust has become one of most importance criterion in order to establish the e-Commerce. Previous study has shown that, there are five most important trust attributes that should be placed in an e-Commerce website. However, consumers may not know about the trust attributes and incapable to search the attributes manually especially for the new consumers. Besides, there are no proper tools available that can help the consumers to search all five trust attributes in an e-Commerce website. In order to overcome the problem, tool that can search the trust attributes should be develop to help the customer to assess the trustworthiness level of an e-Commerce website. Since this study only focus on e-Commerce websites in United Kingdom (UK), United States (US) and Malaysia, data from the websites have been collected to create patterns using regular expressions. Observation has shown that most of trust attributes are located in "Homepage" and "Contact us" pages. The proposed patterns then are applied into the prototype and the practicability of the prototype has been tested. ROC has been used to assign weight to each attributes depending on the rank of the importance of the trust attributes to help the customer to assess and making a decision about the trustworthiness of an e-Commerce website. Precision, recall and f-measure have been used to measure the performance of the prototype and the results show that the system performs almost same result by using manual search.