Design And Implementation Of Index Analysis System For Personalized Recommendation Algorithm Platform

With the rapid development of the Internet industry in today's society,China has entered the era of Web 3.0 in an all-round way,the scale of Chinese netizens and the Internet penetration rate.The scale of Chinese netizens is increasing year by year,accounting for more than half of the total domestic population,reaching 800 million people in June 2018,and the Internet penetration rate is also showing an upward trend.Until June 2018,the Internet penetration rate has increased to 57.7%.As one of the typical representatives of web3.0,the influence of microblog on netizens has been increasing year by year,and it has gradually developed into one of the indispensable social media in the current market of our country.The number of users of micro-blog shows a growing trend.In the process of daily communication,netizens share their lives with the outside world through micro-blog,pay close attention to each other with friends,comment on topics of interest and so on.The huge number of users plus multiple actions brings huge data.But in the face of such a large amount of information,users are only interested in a part of it.How to recommend content that is really interesting to users is the work of recommendation algorithm.At present,the major social media platforms have introduced personalized recommendation function,but how to measure the recommendation effect of recommendation algorithm has become a very important issue.On this basis,this paper realizes the data diversification display function of algorithm effect,which provides data support for algorithmists to design algorithms and observe algorithm effect,and also provides data visualization support for product operators to write reports and improve products.In this paper,factor analysis method is used to score the effect of the algorithm,which provides a certain reference value for users to compare the recommendation effect of different parameter algorithms.Users can also set parameters independently and self-define the effect score,so as to achieve the purpose of observing the effect between the algorithms.The evaluation of algorithm effect can provide users with more accurate reference meaning,and provide users with reference for adjusting algorithm and decision support.