Social media network event propagation key time prediction method, system, and medium

Abstract

The invention belongs to the technical field of online information propagation prediction, and discloses a social media network event propagation key time prediction method, a system, and a medium. The method comprises the steps: carrying out the classification according to different time sequence features of social media network event online information; carrying out smoothing processing on the popularity time sequence with too strong volatility by adopting a Hort linear trend method; recognizing a time interval in which a key node propagated by the preprocessed social media network event occurs; carrying out time window division on the preprocessed time series data, and extracting time series, fluctuation and text emotion features based on online information data; constructing a prediction model training sample, and training a prediction model by adopting an XGBoost algorithm according to the training sample and the number of future time windows; and predicting the occurrence time ofthe key nodes in the event propagation process of the social media network by adopting the trained model. The method can effectively predict the occurrence time of the key node in the event propagation process of the social media network.

Description

technical field [0001] The invention belongs to the technical field of online information dissemination prediction, and in particular relates to a method, system and medium for predicting critical time of social media network event dissemination. Background technique [0002] At present: With the development of the mobile Internet, the influence and speed of social media have greatly surpassed traditional mainstream media. Typical social media such as Twitter, Weibo, and WeChat have become important platforms for public communication and discussion. Social media has endowed users with the autonomy to create content, and the increasingly large user groups and increasingly rich forms of communication have enabled users to publish more and more content, which has brought about an explosive growth in the amount of information. The number of users of the most popular social media has reached 100 million levels, and the amount of data generated every day reaches TB (terabytes) or ...

AI Technical Summary

Problems solved by technology

The disadvantage of this type of method is that it only predicts the popularity of online information at a certain point in the future, but fails to grasp the overall development trend of the social media network event propagation process.

The disadvantage of this type of method is that it only predicts the occurrence time of a single key node in the online information dissemination process, the prediction accuracy is low, and it does not pay attention to other key nodes in the dissemination process.

In addition, most of the current research is to use unilateral features to predict the popularity. How to extract the features that affect the evolution of network event popularity in multiple directions to expand the dimension of input information, and improve the accuracy of key node prediction based on these features is Difficulty of the problem

Images