A new multi-label dataset for Web attacks CAPEC classification using machine learning techniques

There are many datasets for training and evaluating models to detect web attacks, labeling each request as normal or attack. Web attack protection tools must provide additional information on the type of detected, in a clear simple way. This paper presents new multi-label dataset classifying attacks based CAPEC classification, way features extraction ASCII values, evaluation several combinations algorithms. Using extract by computing average sum values characters field that compose request, algorithms (LightGBM CatBoost) classification evaluated, complete system is suffering. The test data used come from SR-BH 2020 dataset. Calculating different make up shows its usefulness numeric encoding feature extraction. allows models, also allowing various undergoing. combination two-phase model with MultiOutputClassifier module scikit-learn library, together CatBoost algorithm superiority criticality scenarios. Experimental results indicate machine learning multi-phase leads improved prediction attacks. Also, use suitable about