Users Based On IPTV Big Data Analysis Recommendation System Design And Implementation

Since 2018,Jiangsu Telecom has been developing IPTV business in Jiangsu area.At present,Jiangsu Telecom has about 5000000 IPTV users.Jiangsu TV Station is the only IPTV program content provider in Jiangsu Province.It is mainly responsible for providing high quality IPTV program content to Jiangsu Telecom.With the complexity and diversity of program content,simple classification and typesetting makes it difficult for users to choose from the TV to their real favorite programs,IPTV display form has been unable to meet the growing user needs.IPTV has the characteristics of interaction.Through this feature,Jiangsu TV Station can obtain the basic data of user behavior recorded by Jiangsu Telecom.This paper defines,classifications and counts the user behavior basic data of IPTV obtained by Jiangsu TV station.Through the detailed segmentation of user behavior,the statistics of program viewing,and the data mining and analysis,the IPTV large data analysis system is constructed.This paper then analyzes the characteristics of IPTV user activities and program types,constructs the general process of recommending programs to users,designs several recommendation models,and completes the combination recommendation algorithm corresponding to the recommendation model.The recommendation model is described and analyzed in detail,including user-based Pearson recommendation,content-based recommendation method and so on.Finally,the video recommendation system of IPTV is designed and implemented.Firstly,the design principle and direction of recommendation system are defined.Secondly,the main parts of recommendation system are designed: large data base platform,large data analysis system and intelligent recommendation engine.Finally,the IPTV recommendation system is designed on the distributed server cluster,and the general flow of intelligent recommendation is explained.Finally,the effect of the recommender system is displayed and tested.This paper implements a recommendation scheme that conforms to the user's viewing habits,and explores a personalized user demand satisfaction method.At the same time,optimize the operation mode of IPTV services in competition,stimulate operators to obtain potential user needs from data algorithm analysis,innovate new directions of user services,and explore new business models.