Predictive intelligence in harmful news identification by BERT-based ensemble learning model with text sentiment analysis
作者：
Highlights：
• A harmful news is defined as those explicit or implicit harmful speech in news text that harms people or affects readers’ perception.
• A harmful news identification dataset is established and conduct a correlation analysis.
• A BERT-based model which applies ensemble learning methods with a text sentiment analysis is proposed to identify harmful news.
• Results show that the F1-score of the proposed model reaches 66.3%, an increase of 7.8% compared with that of the previous approach.
摘要
•A harmful news is defined as those explicit or implicit harmful speech in news text that harms people or affects readers’ perception.•A harmful news identification dataset is established and conduct a correlation analysis.•A BERT-based model which applies ensemble learning methods with a text sentiment analysis is proposed to identify harmful news.•Results show that the F1-score of the proposed model reaches 66.3%, an increase of 7.8% compared with that of the previous approach.
论文关键词：Information disorder,Harmful news analysis,Natural language processing,News sentiment analysis,Ensemble learning,BERT model
论文评审过程：Received 31 October 2021, Revised 24 December 2021, Accepted 8 January 2022, Available online 25 January 2022, Version of Record 25 January 2022.
论文官网地址：https://doi.org/10.1016/j.ipm.2022.102872
