Design And Implementation Of Intelligent Travel Recommendation System For Chongqing City

The rapid development of network information technology has made Internet information technology more closely linked to traditional service industries.The intelligent travel recommendation system uses electronic information technology to classify travel resources and recommend travel resources to users.Aiming at the problems that existing recommendation systems do not have personalized recommendations and the safety emergency treatment is incomplete,this thesis designs and implements a new smart travel recommendation system through a comprehensive analysis of the existing recommendation system architecture,interaction mode,recommendation algorithm and security processing.The prototype system is designed and developed for the thesis of demand analysis.The system mainly consists of personal center,recommendation center,exchange center,itinerary center,and security center.JAVA is selected as the development language and My SQL is used as the database.The system designed the overall development plan of the system software based on the B/S(Browser/Server)architecture,adopting MVC(Model View Controller)based design patterns,and improved code usability and scalability.And adopt a distributed three-tier architecture to achieve separation of interface display,business processing and data support.For personalized recommendation,by analyzing the five commonly used recommendation algorithms and combining the complex factor characteristics of the smart travel recommendation system,this thesis finally uses collaborative filtering algorithms to recommend users.Aiming at the problem that the cold start,sparse matrix of the collaborative filtering algorithm,and the isolated points faced by matrix decomposition lead to inaccurate recommendation results,an M＿CF(M-Collaborative Filtering)recommendation algorithm based on matrix decomposition and item weights is proposed.The algorithm improves the accuracy of algorithm recommendation by averaging the initial score,recommendation results,and item weights.The initial score is the user's rating matrix for the item,the recommendation result is the matrix obtained by decomposing the user's rating matrix,and the decomposition matrix is reconstructed,and the item weight is the user's like level of the item.By considering the mean square error(RMSE)of the measurement index of the recommendation system,it is obvious that the M＿CF algorithm results are better.Aiming at the problem of imperfect emergency hanging of the recommended system,the background of the system combines Baidu maps and historical events related to security,a one-click alarm through a third-party voice cloud platform,and the method of handing over information to emergency contacts or third parties to ensure the safety of users and the smooth handling of emergency events.Finally,this thesis verifies the reliability and effectiveness of the smart travel recommendation system,and performs a functional test and a performance test of the system.The test results show that the smart tourism recommendation system meets the project requirements and achieves the expected goals.