Retrieve information resources made by the machine processing may refer to multiple sources. A personal web as part of information resources in the Internet requires a feature that can be understood by computer machines. Therefore, in this paper an ontology semantic web approach is used to map the resources in a meaningful scheme. In the design of concept, resources on the web are viewed as documents that have some property and ownership. Domain interest or web scope is used to describe a classification of resources that navigate into relevant documents. If instances are completed to the concept, then the ontology file can be loaded and shared as annotation on personal web. This allows computer machine to query multiple ontology from different personal webs that use it.