Research And Implementation Of Mobile Game Recommendation System Based On Html5

With the advent of the Internet era and the innovation of Html5 technology,mobile Html5 games become the game industry "Red Sea".Personalized recommender system can help users find their interest quickly and efficiently.Aiming at the characteristics and development trend of the Html5 game industry,we study the related technologies such as Html5,recommendation algorithm and implement a recommender system.The main works are as follows:(1)According to the requirement analysis,the recommender system is divided into four parts,which are the hall of mobile Html5 games,the background management system of mobile Html5 games,the background service part of the server and game recommendation module.The hall of mobile Html5 games is made to collect the user behavior data,and to provide users with quality services and information recommendation.The background management system of mobile Html5 games is made to manage the data of entire system,to add game,and to modify the settings of the game hall.The background management can also query the game statistics and related users' behavior data.The background service part of the server provides service for the front-end system.The part has the characteristics of high performance and high stability.Game recommendation module provides the recommended game services.It need a strong real-time recommendation with a good recommendation algorithm.(2)We study entry-based collaborative filtering recommendation algorithm.With the similarity measure of the information entropy,we solve the shortcomings of traditional similarity measurement methods and improve the recommend accuracy without changing the data scarcity.This algorithm is validated on the MovieLens dataset and data collected by system.This method is used as the basis of recommendation algorithm of recommender system.(3)We design and implement Html5-based mobile game recommender system,which uses a high real-time collaborative filtering recommendation algorithm for the recommendation of the game,the algorithm consists of data preprocessing,user nearest neighborhood modification and game recommendation.The system has been put into use on third party websites.By analyzing the user behavior data of the system and the income data of the third-party website,the Html5-based mobile game recommender system designed in this paper can recommend the game to users correctly and effectively,which achieve the desired goal.