Collaborative Filtering Recommendation Algorithm Based On Interval Semi-Supervised

Collaborative filtering recommendation algorithm is one of the most commonly used recommendation algorithms.It uses a large number of related data to analyze user behavior similarity and provide personalized recommendation for users.In the era of big data,a large number of data are presented to people.The implementation of personalized recommendation needs to mine the implicit information in the data.LDA topic model is usually used to obtain the topic distribution information of documents.Therefore,many scholars try to apply LDA topic model to collaborative filtering recommendation algorithm,and continue to explore and optimize it.Traditional LDA topic model is unsupervised.In practical text mining applications,the whole data set often needs to deal with a large number of topics,and users are actually interested in only a small part of them.In this case,direct application of LDA model has obvious shortcomings.That is to say,LDA model often generates a lot of local maximum in the process of dealing with a large number of topics,which makes it possible for the model to give many "garbage" topics and eventually generate unstable results.However,in most cases,there are too few related topics in the data set,so a large number of topics need to be used to capture them in an unsupervised way.Based on the LDA topic model,this paper fixes the topic interval corresponding to the relevant topic keywords,and proposes an interval semi-supervised LDA topic model.Interval semi-supervised LDA topic model can effectively avoid the risk of generating "garbage" topic in the process of building traditional LDA topic model,and improve the accuracy of topic distribution calculation.Known collaborative filtering recommendation algorithm can use LDA topic model to calculate the topic distribution of documents,and on this basis calculate the similarity between different documents,so as to get personalized recommendation results.Therefore,this paper applies interval semi-supervised LDA topic model to collaborative filtering recommendation algorithm,which effectively improves the accuracy of algorithm recommendation.This paper uses today's headlines to verify the collaborative filtering recommendation algorithm based on interval semi-supervised LDA.The accuracy of model recommendation is calculated by judging whether the recommendation document belongs to the same category as the current document and whether the content is relevant.The experimental results show that this model can effectively improve the accuracy of user recommendation when users are interested in predefined subject areas.