A common problem facing mobile computing is disconnected operation, or computing in the absence of a network. Hoarding eases disconnected operation by selecting a subset of files for local storage. The purpose of this work focuses on the extraction of Web site access patterns. These patterns permit to establish access probabilities to Web sites, aiming at preselecting and replicating HTML files in a mobile computer. This approach of resource preloading in mobile computer permits the continuity of clients work without caring for the state of their connection to the network. Our system utilizes association rules that are extracted by data mining techniques for determining the set of items that should be hoarded to a mobile computer prior disconnection.