An Intelligent Query System Based on Chinese Short Message Service for Restaurant Recommendation

With the development of modern communication technology, the mobile communication market has grown impressively in China. China mobile and China Unicom have set up the largest mobile communication networks in the world. Concurrently, the number of Chinese short message service (SMS) user has been increased rapidly in recent 10 years. This huge customer infrastructure means immensurable market potential. The SMS-based restaurant recommendation is one of such services with potential market in China. With the promotion of their living level, more and more Chinese people spend their spare time on exciting trips to experience different custom and culture, especially the splendid catering culture in China. Therefore, they are most interested in where they can enjoy their favorite food while traveling in an unfamiliar city, which brings new commercial opportunities. We develop a SMS-based prototype system to help travelers find suitable restaurants. The architecture of the restaurant recommendation system. We choose SMS as the commutation technique between mobile devices and the restaurant recommendation server. As an unnecessarily advanced technology, SMS provides an agile communication way at a lower price (communication service providers charge for sending a short message much less than connecting a call between different cities in China), and is popular among Chinese people, especially among the youth. The most important feature of this system is the ability to deal with the complexities of Chinese language query. A natural language interface of database query is adopted to simplify the user´s interaction with the system. Users can send their special requirements to the restaurant recommendation server in constrained Chinese natural language form. The system then analyzes the words, grammar and semantic content of the message in order to precisely determine the most adequate choice. Moreover, the intelligence of the system is enhanced to meet users´ needs. Before ret- urning the results, the server will carry out a knowledge-based reasoning process to find out the restaurants that qualify for supplying user´s favorite dishes with appropriate prices. A fuzzy method is introduced to solve the problem of understanding uncertain expressions of human languages, which makes the results of reasoning fit users´ requirements more closely. Furthermore, the system will also offer a user the shortest paths to the suitable restaurants if the user´s mobile terminal is equipped with GIS/GPS modules. This work provides a new applicable field for mobile devices and improves the intelligence of their services. Its application will provide new opportunities for both users and providers of recommendation service in tourism industries.