Investigation into the Causes of User Data Breaches Based on Improved Discriminant Coefficient and Gray Correlation Analysis

To accurately identify the primary factors behind user data breach incidents, this study introduces a gray correlation-based method for investigating the causes of user data breaches. From a user's perspective, relevant factors influencing data breach incidents were analyzed and assessed. By adjusting the size of discriminant coefficient based on the dynamic degree of data sequence fluctuations in the dataset, a gray correlation analysis model was established to identify significant influencing factors. A simulation experiment conducted using Kaggle's publicly available "List of Top Data Breaches (2004-2021)" dataset revealed that the most notable factors contributing to data breach incidents include low complexity of user passwords, clicking on unknown links, installing malicious software, leaking sensitive information, failing to back up important data, and not using multi-factor authentication.