Soft computing techniques for product filtering in E-commerce personalisation: A comparison study

In this paper, we compare two soft computing methods used for product filtering in web personalisation for E-commerce. Due to the diversely behaving nature, and the complexity to model the customers´ behaviour using market research methodologies, it is difficult to build a universal model relating the purchasing behaviour mathematical in E-commerce. For this reason, soft computing techniques may be considered as more appropriate in such case. In this study, we have investigated and compared an artificial neural network (ANN) and a fuzzy based method on a particular simulated data set. Initial results indicated that the fuzzy method could be a better choice as there are means to improve the results and human users may understand and modify the model.