Design And Implementation Of Weibo Advertising Recommendation System Based On User Interest

At present,the rapid development of Internet technology has introduced human beings into the era of big data.In the ocean of data,it is difficult for information consumers to effectively find interesting and valuable information,and information producers cannot deliver information to specific users,this is the "information overload" problem in the era of big data.Weibo(formerly known as Sina Weibo)is one of the best-developed social products since the web 2.0 era.It currently has about 489 million monthly active users.For Weibo,its mobile and web terminals generate huge daily data.The amount of information makes the "information overload" problem on the platform very serious.To solve this problem,this article uses recommendation technology to build a recommendation system.It is worth mentioning that "social media marketing",the most popular Internet marketing concept bred by Internet technology and e-commerce companies,is widely used by some companies to promote products.They use Weibo accounts to publish related products.Blog posts or purchase of Weibo platform advertising functions to promote the published blog posts to the platform users,but the "information overload" problem of the Weibo platform makes it easy for users to miss the Weibo posted on the Weibo homepage by interested companies.The function of Weibo advertisement often appears that the promotion blog post does not match the user's interest.Therefore,this article takes Weibo as the research object and develops a Weibo advertisement recommendation system based on user interests.This can both solve the problem of "information overload" and recommend Weibo ads for users based on their interests.The main task of this paper is to design and implement a Weibo advertisement recommendation system based on user interests.To this end,this paper works from the following aspects:(1)Collect microblog data.There are many ways to obtain Weibo data.This article uses web crawler technology to simulate user login,design crawling strategies,extract Weibo data and store the data in the database;(2)Model the user interest based on the Weibo data and feature extraction of Weibo ads.This article uses the LDA topic model to mine user interests from three aspects,including user original blog posts,user social relationships,and user interaction behaviors,and uses the HowNet lexical semantic tendency calculation method to merge synonyms to obtain the final user interest model.(3)Use TFIDF algorithm to extract the characteristics of Weibo ads.(4)On the basis of constructing a user interest model to obtain user interest characteristics,research on the personalized recommendation technology of Weibo ads.This paper proposes personalized recommendation algorithms based on the LDA topic model and improved collaborative filtering to recommend Weibo ads for users.(5)Finally,this article uses the Python language and Django website development framework and MySQL database to develop a Weibo advertisement recommendation system based on user interests.Based on the above work,this article develops a Weibo advertisement recommendation system based on user interests.Ordinary users and administrator users log in to the system through their accounts.Ordinary users can view the list of Weibo advertisements recommended by the system,as well as user profiles and post information.The administrator can start a data collection program,build a user interest model,extract advertising features,and personalize Weibo ad recommendations.