A novel semi-supervised self-training method based on resampling for Twitter fake account identification.
Ziming ZengTingting LiShouqiang SunJingjing SunJie YinPublished in: Data Technol. Appl. (2022)
Keyphrases
- semi supervised
- experimental evaluation
- pairwise
- segmentation method
- cost function
- high accuracy
- detection method
- social networks
- similarity measure
- computational cost
- neural network
- dynamic programming
- synthetic data
- multi class
- high precision
- classification method
- support vector machine svm
- probabilistic model
- significant improvement
- computational complexity
- learning algorithm