ERP-Guard: An Advanced NLP Based System for the Early Prediction of Cyberattacks by Analyzing Severity of Risk

Abstract

The principal idea of this study is to discuss the use of cutting-edge methods to anticipate and evaluate cyberthreats, this paper seeks to strengthen the security of Enterprise Resource Planning (ERP) systems. we designed ERP log files with cutting-edge Natural Language Processing (NLP), the Common Vulnerability Scoring System (CVSS), and a fine-tuned XLNet model trained on the National Vulnerability Database (NVD). Accurately estimating risk levels and quantifying vulnerabilities in ERP records are the main goals. Methodologically, the study analyzes ERP log data by utilizing the bidirectional context comprehension of the XLNet model. offers a thorough understanding of vulnerabilities found inside ERP logs, boosting proactive protection systems against cyber threats. The main result employs CVSS ratings, which range from 0 to 10, to identify the risk; higher scores indicate greater severity. By providing a thorough understanding of vulnerabilities detected in ERP records, this risk evaluation tools solution increase proactive defences against cyberattacks.