Method,Accuracy,F1 (macro),AUC-ROC,Source
HyperModel (Stacking Ensemble),0.9888435875582583,0.9875919887801082,0.9989987226653723,Ours
CharCNN,0.98718509816568,0.9857623867375606,0.9985083567225014,Ours
LSTM (Bahnsen 2018),0.987,,,Literature
CNN-BiLSTM,0.9866936939012124,0.9852331088044274,0.998483685483307,Ours
URLNet CharCNN (Le 2018),0.985,,0.986,Literature
CharBiLSTM,0.9844439837529465,0.9827525986682384,0.997992440144913,Ours
XGBoost (Optuna-tuned),0.9815876964657284,0.9794700103530696,0.9973786670930095,Ours
LightGBM (Optuna-tuned),0.9813957416749207,0.9793213647507326,0.9971396658702955,Ours
TF-IDF + LR,0.9803438294212947,0.9780726993573674,0.9971270195370545,Ours
XGBoost (GPU),0.980174909205384,0.9778762050700486,0.996815861156588,Ours
RF on URL features (Sahingoz 2019),0.9798,0.9793,,Literature
LightGBM (CPU),0.9781862575726165,0.9757796254760449,0.9965268674959424,Ours
Stacking CNN+LSTM+GRU (Nguyen 2021),0.975,0.972,,Literature
XGBoost (Rao & Pais 2019),0.9685,,0.993,Literature
CharCNN (Aljofey 2022),0.9675,0.9671,0.991,Literature
Hybrid DNN+LSTM (Ozcan 2021),0.964,,0.973,Literature