This research proposes a novel method for forecasting possible attacker approaches in information systems by utilizing both association rule mining and an adapted FP-Growth algorithm. To facilitate the achievement of this objective, the study first established operational boundaries from both business and technical perspectives. A comprehensive examination of association rule generation techniques was performed, with particular focus on adapting the FP-Growth technique to the cybersecurity context. A dataset was compiled from open-source information security threats, which served as the basis for experimentation. The performance of the proposed algorithm was then thoroughly assessed, with the model parameters adjusted to maximize prediction accuracy. Verification of the findings confirmed that the enhanced FP-Growth approach effectively identifies potential attack vectors, thus providing valuable insights for information system protection and highlighting the algorithm’s applicability for practical cybersecurity defense.