Volumes and types of all kinds of electronically available data have increased dramatically in recent years as a result of the Internet being primary means of communication and collaboration among people of various educational background and diversity. Current search engines do not reach most of these information sources on the Internet. Although, search engines access static URL links easily, because of the decentralized nature of the Internet, we face a chaotic repository of all types of information. Therefore, it is becoming more and more difficult to locate resources of interest in the web. In response to user's queries, information discovery tools such as search engines produce a list which, sometimes, includes some unnecessary documents, and occasionally may leave out necessary and important ones for a querying user. In this thesis, we design a model for web information discovery tools. Users and documents are defined in five aspects by metadata. Conceptual relations are set between user and documents. While searching a list of results was obtained by using the relationship between users and documents. This result list was ranked by using PageRank algorithm.