AI-Driven Log Forensics System For Intelligent Threat Detection
DOI:
https://doi.org/10.29284/y0tmva88Keywords:
SVM, NER , BERT , LSTM, DNN , TF-IDFAbstract
Fast, growing digital worlds plus smarter cyber attacks have made studying log files key to staying safe online. Because systems record what users do, how apps behave, and when something odd happens, those records help spot break, or misuse from inside. Even so, too much data coming too fast in too many formats overwhelms old, school tools, leading to slow responses, missed clues, or wrong alarms. To fix this mess, a new method uses artificial intelligence to gather, clean up, and examine different kinds of logs without constant human help. Hidden links and odd shifts in behaviour get spotted by smart pattern, finding tools built into the system. Instead of just following rules, it learns what usual activity looks like, then notices when something unfamiliar creeps in, even if it has never shown up before. While watching live data streams, it adjusts itself based on new threats, shifting its understanding without waiting for updates. Predictions about future risks come from trends pulled straight out of current signals, not guesses. Speed matters, so alerts fire quickly once a meaningful change appears.
References
[1] A new method uses artificial thinking to catch odd behavior right away. Instead of waiting, alerts happen during attacks. This approach helps teams react before damage spreads. Work by Ogochukwu Susan Ndibe shows how timing changes outcomes. By 2025, such tools could reshape defense strategies.
[2] Simran Keshari, Mitali Srivastava, Cyber Attack Prediction: Moving from Traditional Machine Learning to Generative Artificial Intelligence, 2025.
[3] Log Poisoning Attacks in IoT: Methodologies, Evasion, Detection, Mitigation, and Criticality Analysis, 2024.
[4] Meta Log: Generalizable Cross, System Anomaly.
[5] Detection from Logs with Meta, Learning, 2024.The Evolution Of Automated Log Analysis.
[6] Anomaly detection Algorithms for Large Scale Real, Time Log Data Analysis, 2022.
[7] Deep Learning for Anomaly Detection in Log Data: A Survey, M. Landauer et al., 2023.
[8] Deep Learning for Anomaly Detection in Log Data: A Survey (arXiv preprint), Landauer et al., 2022.
[9] A Study of Machine Learning Techniques for Log, Based Anomaly Detection, S. Ali et al., 2025.
[10] Impact of Log Parsing on Deep Learning, Based Anomaly Detection, Z.A. Khan et al., 2024.
[11] Log Anomaly Detection via Evidential Deep Learning, Y. Duan et al., 2024.
[12] Host, based intrusion detection with multi, data source and deep learning, R.H. Hwang et al., 2023.
[13] Machine learning, based intrusion detection framework for infrastructure, V. Kantharaju et al., 2024.
[14] Evaluating machine learning, based intrusion detection with Explainable AI, V.Z. Mohale et al., 2025.
[15] Determining the best features in log files for anomaly detection through the use of optimization techniques, Shivaprakash Ranga, 2025.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
