In view of the defects of the traditional user modeling technology in personalized recommendation system, this paper introduces an ontology-based user modeling method. In order to retrieve the most interested goods from a large amount of information for the user, a recommendation model is designed which combines the ontology technology and tourism information. This paper mainly studies the knowledge representation in tourism field and ontology-based model building method. This paper describes the implementation procedure of how to use the ontology editor tool Protégé to construct OntoTRec, and points out the procedure and method how to realize it. The feasibility and effectiveness of recommendation algorithm proposed in this paper have been verified on real-world data sets, and it achieved higher prediction accuracy compared with existing recommendation algorithms