图/表 详细信息

基于动态增强图注意力网络的突发事件预测
仲兆满, 崔心如, 张渝, 吕慧慧, 樊继冬
南京大学学报(自然科学版), 2025, 61(1): 94-104.   DOI: 10.13232/j.cnki.jnju.2025.01.009

ThailandEgyptIndiaRussia
F1RecPrecF1RecPrecF1RecPrecF1RecPrec

Non⁃

Temp

LR (Count)*0.77010.71290.83720.79450.74680.84880.61820.55890.69160.73890.72050.7582
LR (word TF⁃IDF)*0.71510.63370.82050.77950.75110.81020.54300.43350.72660.70480.68940.7208
LR (N⁃GramTF⁃IDF)*0.72930.65350.82500.76100.70390.82830.55150.44110.73550.71430.71430.7143
GCN*0.76130.75800.76630.84910.81610.87870.65330.62710.68530.78400.82620.7469
GAT0.76770.75320.78280.83380.82580.84190.64670.61780.67850.78300.81570.7529
TempnMLL*0.73040.66140.81550.72340.79690.66230.62770.71930.55670.75950.76920.7500
GCN+GRU*0.78250.76860.79990.85000.82500.87750.65470.62150.69630.78660.80870.7677
GCN+LSTM*0.78130.77020.79380.85070.82710.87660.64930.61370.69140.78580.79140.7812
GCN+RNN*0.75660.75530.75850.85100.82040.88510.64160.60160.68920.78680.80880.7667
DynamicGCN*0.79700.77340.82480.86170.82850.89840.66920.62750.71960.80400.79880.8101
EvolveGCN0.79110.75690.82850.84940.81270.88930.66310.62880.70210.73650.76700.7081
HGT0.80450.77220.83960.84540.79910.89770.67470.65560.69540.75930.78680.7337
EGAT+GRU0.75420.69660.82210.82450.80370.84640.68020.66340.69790.80690.83960.7768
EGAT+LSTM0.79130.75170.83530.83990.82420.85640.65970.64040.68030.79410.80680.7815
EGAT+RNN0.80030.77330.82920.77230.79740.74870.65100.68980.61640.77850.82990.7325
DEGAT0.83580.80770.86600.86190.82350.89850.68930.67260.70690.85280.86780.8384
表3 模型性能对比
本文的其它图/表