Design Of Spots Recommendation System Based On Unity3D

With the development of hardware technology and digital application,the traditional spots recommendation system,which introducs tourist attractions mainly by integrating the information of text and pictures,is hard to satisfy users preference.The fundamental reason is that spots recommended seem to be all the same for all users,which is difficult to arouse users interests.Fortunately,a new model,which combining personalized recommendation and virtual reality technology has been widely accepted by researchers,because virtual reality technology offers users the most authentic experience and personalized recommendations can meet the needs of different users.In this paper,the recommender system framework is constructed based on the scenic spots around Xiamen University,incorporating 3D virtual technology and personalized spots recommendation algorithm.The former can make users feel immersive,and the latter is able to provide personalized service for each user.This system mainly consists of two parts,which are platform construction and algorithm design.We first make three dimensional model of spots,containing modeling the buildings and roads using SketchUp and modeling character using 3Dmax.All the codes are implemented in C#and all the models are conducted on Unity3D.The modeling process is divided into five steps,that is data acquisition,processing,modeling,texture rendering,model optimization and export.fbx files.Unity3D is mainly used for display scenic spots,user interaction and recommend different routes according to users’ preference and constraint conditions.Users can evaluate each spot to optimize the recommended route at the same time.We use two recommendation algorithms,one is popular scenic spots recommendation algorithm,the otherone is personalized recommendation algorithm.The popular scenic spots recommendation algorithm is derived from greedy algorithm,it obtains the ranking of spots,which are used to design the routes,through the questionnaire.Personalized recommendation algorithm,which is based on A*algorithm,recommends personalized route according to the users’ preferences and constraint conditions in order to achieve users’ maximum satisfaction.The experiments focus on analyzing the running efficiency of the algorithm and comparing the performance of routes based on users’ perference and condition.