Authority And Hub

Television ratings plays an important role in the development of television media.However,on the one hand,the traditional ratings surveys are strictly restricted to the sample size,on the other hand,as the TV ratings fraud appeared in recent years,people are increasingly skeptical about the independence and credibility of the TV ratings data.With the advent of the multi-screen era,the TV industry's works and the audience's television viewing behaviors have changed.There are serious flaws in the ratings evaluation.The practical problems faced by TV media practitioners are how to solve the current predicament.The set-top box of cable TV has recorded a large amount of audiences' behavior data,which provides favorable conditions for solving this problem.Under this circumstance,this paper studies ratings analysis based on the HITS algorithm and explores the new ways of evaluating ratings.Firstly,we introduce the research background,purpose,significance and research methods of this paper.Secondly,we summarizes the research in ratings and ratings surveys and its related area.Based on the analysis ratings of the research status and measurement methods,the paper points out the problems existing in the ratings.In the end,we propose the research content:the television ratings analysis based on the HITS algorithm.Because the HITS algorithm can calculate the score of programs and users in the bipartite graph by several times of iterative computation.So it can evaluate the hub of the user and the authority of the program.The study use Python to clean data and extract information from yangzhou television data provided by the Zhongshu company,and Hits algorithm is applied to estimate the ratings.To test the rationality of the HITS algorithms,the study calculates the traditional ratings based on viewers number or viewing time at the same time.The data analysis result shows that:On the one hand,there is a strong correlation between the HITS algorithm and the raditional ratings algorithm which the correlation coefficient is up to 0.96,so it is feasible to use HITS calculate ratings.On the other hand,there are obvious differences between the two.This paper argues that the HITS algorithm calculate ratings has a greater advantage,and can vary depending on the ratings,and it attempts to evaluate ratings from different angles.By using the HITS,we can infer that television viewer watching behavior,to provide on the basis of the television program selection and programming for the television station,while providing advertising on-air and evaluation effectively for advertising operators;This algorithm can be used to analyze the ratings of video website programs.