Research And Implementation Of Personalized Travel Recommendation System Based On Domain Ontology

With the sustainable improvement of living level, tourism takes an important part in people’s life. People get abundant tourism resources by internet because of the internet technology boom. But in face of such a rich tourism resources, people lack of effective information filter measures to gain information they need. It leads to the problem of "information overload". The appearance of recommendation system can effectively solve the information overload problem, it can dig the user interest preference actively, push the interesting information to the user.Through the detailed investigation to the current situation of the domestic and foreign tourism recommendation systems, the author finds that most of the current foreign tourism recommendation systems seldom deeply comprehend user and item. And these recommendation algorithms used some limited user and item description files, did not make full use of the user’s interaction history and the item characteristic attributes. Especially, domestic tourism websites just provide the basic tourism information query service for user, and lack of personalized recommendation service.Focus on the existing problems of the tourism recommendation system above, this paper proposes the personalized tourism recommendation system based on domain ontology. This paper has accomplished the following tasks:① Through the detailed investigation to the current situation of the domestic and foreign tourism recommendation systems, the paper aimed at the existing problems of the tourism recommendation system, put forward the personalized tourism recommendation system based on domain ontology.② The related content of ontology theory and the composition of domain ontology are analyzed. The influence factors of concepts of ontology tree similarity calculation are explored, and the paper analyzed some concept similarity computation methods based on ontology. Finally, the paper introduces personalized recommendation algorithm and system model.③ Ontology construction methods are researched, the paper analyzed the tourism information by seven step method and built the tourism domain ontology by the software of protégé. Making use of tourism domain ontology and the existing algorithm, the paper combined a concept similarity algorithm based on semantic distance with Wu and Palmer algorithm and proposed the concept similarity algorithm based on tourism domain ontology. At last, the paper used the concept similarity algorithm based on tourism domain ontology improved the collaborative filtering algorithm based on ontology and compared the proposed algorithm with original algorithm and traditional collaborative filtering algorithm through the experiment.④ Based on the design target of the personalized tourism recommendation system, the paper proposed the framework of the personalized tourism recommendation system. The paper detailed divided the personalized tourism recommendation function module based on the system framework. Finally, the paper designed the database of the personalized tourism recommendation system.⑤ Based on the system design, the paper implemented the personalized tourism recommendation system based on domain ontology by the technology of Java Web and displayed the effect of the key function. The effect of implementation suggests that the system can solve the problem which most tourism sites lack of personalized recommendation service well and the proposed algorithm make use of the item characteristic attributes well.