News life cycle prediction method and system based on complex network structure

Abstract

The invention discloses a news life cycle prediction method based on a complex network structure. The method comprises the following steps: S1, acquiring and cleaning a data set; S2, constructing a complex network of news and comments; s3, generating features of the news comment tree based on a hierarchical clustering algorithm; s4, extracting the news theme through an LDA theme; s5, combining the news text features with the structural features of the complex network, and dividing the news into a longer life cycle, a medium life cycle and a shorter life cycle through a random forest model; inputting the obtained features into a SVM regression model again to predict the life cycle of the news. The invention further discloses a system for implementing the news life cycle prediction method based on the complex network structure. According to the method, the text features of the news and the structural features of the news comment network can be extracted from the data set, the prediction accuracy is improved, and meanwhile, the overall robustness and universality of the model are improved.

Description

technical field [0001] The invention relates to data mining, machine learning and graph embedding technology, in particular to a method and system for calculating and processing complex network structure features and predicting news life cycle. Background technique [0002] In recent years, with the popularization of the Internet and the rapid development of information technology, traditional books and paper media have been gradually replaced by Internet products. As a platform for users to exchange information, the Internet occupies an important position in users' work and life. Users can obtain various information through the Internet. How to predict the life cycle of information is particularly important. At the present stage, there are mainly three types of life cycle prediction methods: empirical discrimination method, mathematical model method and joint method. The most commonly used product life cycle prediction methods in the early stage mostly focused on qualitat...

AI Technical Summary

Problems solved by technology

However, these traditional methods to predict the life cycle have certain errors in the results; at the same time, due to the variability of news, these methods are not suitable for predicting the life cycle of news

[0003] Chinese patent CN202110395553 discloses a method and system for predicting the life cycle of parking lot equipment, which has relatively large limitations; Chinese patent CN201910062267.6 discloses a method combining complex networks and machine learning, which is applied to the prediction of tumor driver genes ; while in terms of life cycle prediction, there are few applications of complex networks

Images