Od-ids2022: generating a new offensive defensive intrusion detection dataset for machine learning-based attack classification

Od-ids2022: generating a new offensive defensive intrusion detection dataset for machine learning-based attack classification

연구 분야: Analysis

학회: International Journal of Information Technology

In network defence, intrusion detection is crucial to identify malicious activities such as attacks, intrusions, and malware. Intrusion Detection Systems (IDSs) are mandatory for protecting critical networks against intrusive activities. Despite advancements in IDS research, a significant challenge is finding comprehensive and valid datasets to evaluate proposed techniques. To address this, researchers have proposed eleven desirable characteristics for IDS datasets, including Attack Diversity, Complete Traffic, and Metadata. However, existing IDS datasets need to meet these characteristics. To overcome this, we propose the Offensive Defensive-Intrusion Detection System (OD-IDS2022) dataset, a comprehensive and empirical IDS dataset with the latest attacks. The dataset contains benign and twenty-eight common attacks and satisfies all eleven desirable characteristics. We evaluate the performance of four state-of-the-art Machine Learning algorithms (Random Forest, Decision Tree, Naive Bayes, and Support Vector Machine (SVM)) on OD-IDS2022 and find that SVM provides the highest prediction accuracy for both the training and validation samples.