The continued growth of the World Wide Web makes the need for retrieval of relevant information for a userís query increasingly important. Current search engines provide the user many Web pages, but with varying levels of relevancy. In this research, an architecture is presented for the development of an intelligent agent methodology to automate the processing of a userís query, while taking into account the queryís context. Four sample queries are processed simulating the methodology of the agent. The queries differ on two dimensions: the amount of clarity of the domain and the number of related terms. The results of the queries are compared to those obtained from the Google search engine. The comparisons show that applying the intelligent agent methodology to the queries produces significantly fewer Web pages that are equally or more relevant to the user.