Lifecycle research of social media rumor refutation effectiveness based on machine learning and visualization technology
作者：
Highlights：
• Studying the social media rumor refutation effectiveness lifecycle (SMRREL) from three important aspects (i.e., lifespan, peak value, and distribution).
• Exploring possible factors affecting the lifespan and peak value of SMRREL through regression models in the machine learning field (e.g., XGBoostRegressor, LGBMRegressor, CatBoostRegressor, and other ensemble algorithms) and SHapley Additive Explanations method.
• Summarizing distributions of SMRREL with the help of the K-shape clustering algorithm.
• Giving relevant decision-making suggestions to enhance the persistence and intensity of rumor refutation effectiveness.
摘要
•Studying the social media rumor refutation effectiveness lifecycle (SMRREL) from three important aspects (i.e., lifespan, peak value, and distribution).•Exploring possible factors affecting the lifespan and peak value of SMRREL through regression models in the machine learning field (e.g., XGBoostRegressor, LGBMRegressor, CatBoostRegressor, and other ensemble algorithms) and SHapley Additive Explanations method.•Summarizing distributions of SMRREL with the help of the K-shape clustering algorithm.•Giving relevant decision-making suggestions to enhance the persistence and intensity of rumor refutation effectiveness.
论文关键词：Rumor refutation effectiveness,Social media,Lifecycle,XGBoostRegressor
论文评审过程：Received 28 April 2022, Revised 10 July 2022, Accepted 28 August 2022, Available online 20 September 2022, Version of Record 20 September 2022.
论文官网地址：https://doi.org/10.1016/j.ipm.2022.103077
