Fake news spreaders profiling using N-grams of various types and SHAP-based feature selection

Mathematics

• N-gram 100%
• Profiling 84%
• Feature Selection 79%
• Machine Learning 65%
• Transfer Learning 56%
• Learning 54%
• Text Classification 51%
• Classifier 50%
• Natural Language 37%
• Voting 37%
• Engineering 24%
• Random Forest 21%
• Language 21%
• Support Vector Machine 18%
• Radial Basis Function 17%
• Logistic Regression 16%
• Efficacy 15%
• Syntax 14%
• Character 11%
• kernel 10%
• Model 9%
• Performance 8%

Engineering & Materials Science

• Spreaders 79%
• Feature extraction 45%
• Machine learning 43%
• Transfer learning 35%
• Classifiers 31%
• Deep learning 26%
• Logistic regression 16%
• Random forests 15%
• Syntactics 15%
• Support vector machines 11%