Figure 5
The horizontal axis lists “Debunking Climate”, “Debunking Ukraine”, “L I A R”, “FakeNewsNet”, “Emergent”, “Climate-Fact”, “Multi-Fake-DetectIVE”, “NewsPolYML”, “M M C F N D”, and “M F N D”. The vertical axis ranges from 0 to 1.0 in increments of 0.1. The legend is positioned at the bottom center below the graph and lists four models: “L R”, “R F”, “BiLSTM”, and “B E R T”, each represented by a distinct colored bar. For “Debunking Climate”: L R: 0.83, R F: 0.86, BiLSTM: 0.91, B E R T: 0.95. For “Debunking Ukraine”: L R: 0.82, R F: 0.85, BiLSTM: 0.90, B E R T: 0.94. For “L I A R”: L R: 0.64, R F: 0.68, BiLSTM: 0.72, B E R T: 0.77. For “FakeNewsNet”: L R: 0.70, R F: 0.73, BiLSTM: 0.76, B E R T: 0.81. For “Emergent”: L R: 0.66, R F: 0.69, BiLSTM: 0.73, B E R T: 0.78. For “Climate-Fact”: L R: 0.80, R F: 0.83, BiLSTM: 0.87, B E R T: 0.91. For “Multi-Fake-DetectIVE”: L R: 0.81, R F: 0.84, BiLSTM: 0.88, B E R T: 0.92. For “NewsPolYML”: L R: 0.73, R F: 0.76, BiLSTM: 0.80, B E R T: 0.85. For “M M C F N D”: L R: 0.70, R F: 0.74, BiLSTM: 0.78, B E R T: 0.83. For “M F N D”: L R: 0.75, R F: 0.78, BiLSTM: 0.82, B E R T: 0.87.Precision scores resulting from the validation