Dataset for "A Comparative Study of Different Deep Learning Techniques for Software Vulnerability Detection"

datasets Contains train, validation and test splits from 3 different datasets: * BigVul* MegaVul* ProjectKB results test_cases_and_predictions.csv Contains test case level data like CWE-ID, OWASP top 10 category, MVC category, fix size, fix type and also predictions from all three methods.Interpretation of ground_truth and method columns: * 0 means secure* 1 means vulnerable metrics_GNN_LLM_CodeBERT_combined_methods.csv Contains binary classification metrics for the GNN-based, the LLM-based, the CodeBERT-based and combined methods. metrics_per_cwe_GNN, metrics_per_cwe_LLM, metrics_per_cwe_CodeBERT They contain binary classification metrics for the given method separately for each CWE-ID. codebert Contains the code for fine-tuning the CodeBERT-based classifier.