对症下"谣"——谣言评级模型推动差异化管理
Rumor Rating Model to Promote Differentiated Management
刘博金 1张心怡 1纪伯庸 1王靖萱 1常亚乾1
作者信息
- 1. 河北地质大学 管理学院,石家庄 050031
- 折叠
摘要
本研究提出了一种谣言分级的策略.通过深入的调研和数据分析,收集并处理了大量的谣言数据,提取出关键特征,并进行高质量的标注.基于这些数据建立了初步的谣言评级模型,并利用主动学习算法提高了模型的泛化能力.随着数据库的不断扩大,模型的性能将进一步提升.谣言分级具有广泛的应用价值,建议建立谣言等级治理体系,以实现差异化的管理.
Abstract
This study proposes a strategy of rumor grading.Through in-depth research and data analysis,a large amount of rumor data was collected and processed,key features were extracted,and high-quality labeling was performed.Based on these data,a preliminary rumor rating model is established,and the generalization ability of the model is improved using active learning algorithms.With the continuous expansion of the database,the performance of the model will be further improved.Rumor rating has a wide range of application value,and it is recommended to establish a rumor rating governance system to achieve differentiated management.
关键词
谣言评级模型/谣言分级/主动学习算法/机器学习Key words
rumor rating model/rumor grading/active learning algorithm/machine learning引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据