| In recent years, with the emergence of intelligent mobile terminal and the mobile Internet and the popularization, the concept of "O2O" more and more widely. Not only that, but from offline to online more and more, from the reality to the virtual combining site, this combination has been successfully applied in electricity, group, enterprise management, and other industries, in the great game industry users, through LBS can provide mobile terminal users of the advantage of geographical location information, virtual reality games based on LBS have also been more and more attention. This kind of game with the participation of reality, can provide users with more interesting and improve the user experience, if the design is reasonable, will also participate in the games bring real benefits.This thesis mainly introduces three parts.1) The same type of game for the research and analysis, and optimized the game framework, the game is realized by using the Django back end, and build on IOS game front end, such a framework design improves the development of convenience and friendly user experience;2) Combined with the existing type of game, the game model and function are analyzed, and formed a kind of based on LBS O2O puzzle game framework of requirements, namely through LBS for the user’s location information, and through the map as a navigation, search for related puzzle, realize offline solutions, online check, and according to the players habits problem solving history information analysis, recommend related puzzle, system will implement the players score ranking, achieve the purpose of incentive and competitive, improve players’ participation;3) To the existing recommender systems are studied and summarized, and compared the mobile recommender system and the similarities and differences of traditional recommendation system, combined with the mobile recommender system application in LBS, put forward the suitable for LBS recommended strategy, based on the location filter with user collaborative filtering recommendation method based on SVD decomposition, implements the O2O games recommendation system based on LBS.In this thesis, the design and implementation of a kind of based on LBS 020 puzzle game framework has been applied to a company’s business activities, system operation performance is stable. This model effectively promote the user operators to provide integration and planning of the game environment, also promoted the cooperation of the dealer sales service. this thesis designed and implemented the recommendation system is innovative practice in such applications, and for the application in the field of 020 provides beneficial exploration experience. |