Developing an adaptive search engine for E-commerce using a web mining approach

This paper discusses current work using an adaptive learning algorithm to dynamically create the content of an e-commerce search engine so that the implicit knowledge extracted by web text mining module can be provided in the B2B portal. In this work, we develop an algorithmic approach for automatically discovering implicit customer knowledge from Internet by means of a Web mining method. Using a variation of the automatic thesaurus generation techniques namely the Self-Organizing Maps (SOM) neural net, we have conducted several experiments in a specific domain in which we created a functional thesaurus of numerous supplier and product-specific terms. Further, we apply such thesaurus into a topic-hierarchy based text database, as an organized text source of a search engine for a novel B2B E-Commerce Portal.