Using AI for Predictive Debugging in Runtime Environments
DOI:
https://doi.org/10.29284/8d4tj008Keywords:
Artificial Intelligence; Debugging; Error Prediction; Machine Learning; Runtime Environment.Abstract
Runtime failure debugging for advanced software systems is still a time-consuming and error- ridden process, particularly when using conventional rule-based approaches. To counter this, the paper introduces a predictive debugging system based on Hugging Face Transformers, namely the CodeT5 model to detect anomalies and make failure predictions based on sequences of software logs. The approach entails structured preprocessing, temporal log sequencing, and semantic token embedding for enhanced analysis. The data used is obtained from Stack Overflow Logs in Kaggle. ElasticSearch is used for scalable storage of logs, and Kibana is used for real-time visualization and alerting. Experimental outcomes indicate that the designed model has an accuracy over 98%, precision, recall, and F1-scores over 0.97 for both runtime and non-runtime error classes. The confusion matrix validates negligible misclassifications, and the ROC curve results in an AUC of 0.99, which indicates nearly perfect classification performance. Attention heatmaps improve explainability by making salient failure tokens stand out. Overall, the system provides accurate, scalable, and interpretable predictive debugging in intricate runtime environments.
References
Adepoju AH, Austin-Gabriel B, Hamza O, Collins A (2022). Advancing monitoring and alert systems: A proactive approach to improving reliability in complex data ecosystems. IRE Journals, 5(11): 281–282.
Almodovar C, Sabrina F, Karimi S, Azad S (2024). LogFiT: Log anomaly detection using fine-tuned language models. IEEE Transactions on Network and Service Management, 21(2): 1715–1723. Publisher: IEEE.
Atemkeng M, Hamlomo S, Welman B, Oyetunji N, Ataei P, Fendji JLK (2024). Ethics of software programming with generative AI: is programming without generative AI always radical? arXiv preprint arXiv:2408.10554.
Fortier P (2024). Programming language abstractions for the Internet of Things era. PhD Thesis, INSA de Lyon.
Guan W, Cao J, Qian S, Gao J, Ouyang C (2024). Logllm: Log-based anomaly detection using large language models. arXiv preprint arXiv:2411.08561.
Harjunpää N (2023). Log management system technologies and methods for near real-time fault analysis systems: An exploration of log shipping and storage.
Hrusto A (2024). Enhancing DevOps with Autonomous Monitors: A Proactive Approach to Failure Detection. Lund University.
Kumar S (2023). Reviewing software testing models and optimization techniques: an analysis of efficiency and advancement needs. Journal of Computers, Mechanical and Management, 2(1): 32–46.
Lee Y, Kim J, Kang P (2023). Lanobert: System log anomaly detection based on bert masked language model. Applied Soft Computing, 146: 110689. Publisher: Elsevier.
Li X, Huo Y, Mao C, Shan S, Su Y, Li D, et al. (2025). AnomalyGen: An Automated Seman- tic Log Sequence Generation Framework with LLM for Anomaly Detection. arXiv preprint arXiv:2504.12250.
Lim YF, Zhu J, Pang G (2025). Adapting Large Language Models for Parameter-Efficient Log Anomaly Detection. In: Pacific-Asia Conference on Knowledge Discovery and Data Mining, 325–337. Springer.
Qi J, Huang S, Luan Z, Yang S, Fung C, Yang H, et al. (2023). Loggpt: Exploring chatgpt for log-based anomaly detection. In: 2023 IEEE International Conference on High Performance Computing & Communications, Data Science & Systems, Smart City & Dependability in Sen- sor, Cloud & Big Data Systems & Application (HPCC/DSS/SmartCity/DependSys), 273–280. IEEE.
Ugwueze V (2024). Cloud Native Application Development: Best Practices and Challenges.
International Journal of Research Publication and Reviews, 5(12): 2399–2412.
Yang Z, Harris IG (2025). LogLLaMA: Transformer-based log anomaly detection with LLaMA.
arXiv preprint arXiv:2503.14849.
Zanbouri K, Noor-A-Rahim M, John J, Sreenan CJ, Poor HV, Pesch D (2024). A comprehensive survey of wireless time-sensitive networking (TSN): Architecture, technologies, applications, and open issues. IEEE Communications Surveys & Tutorials. Publisher: IEEE.
Zhang L, Jia T, Jia M, Wu Y, Liu A, Yang Y, et al. (2024). A survey of aiops for failure management in the era of large language models. arXiv preprint arXiv:2406.11213.
Zhao C, Ma M, Zhong Z, Zhang S, Tan Z, Xiong X, et al. (2023). Robust multimodal failure detection for microservice systems. In: Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 5639–5649.
Zhou Y, Chen Y, Rao X, Zhou Y, Li Y, Hu C (2024). Leveraging large language models and BERT for log parsing and anomaly detection. Mathematics.
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.
