Log-based anomaly detection with deep learning: how far are we?
Open accessThe DOI matches its Crossref record; 57 of the 57 references it cites were found in a registry or an index.
Abstract
Software-intensive systems produce logs for troubleshooting purposes. Recently, many deep learning models have been proposed to automatically detect system anomalies based on log data. These models typically claim very high detection accuracy. For example, most models report an F-measure greater than 0.9 on the commonly-used HDFS dataset. To achieve a profound understanding of how far we are from solving the problem of log-based anomaly detection, in this paper, we conduct an in-depth analysis of five state-of-the-art deep learning-based models for detecting system anomalies on four public log datasets. Our experiments focus on several aspects of model evaluation, including training data selection, data grouping, class distribution, data noise, and early detection ability. Our results point out that all these aspects have significant impact on the evaluation, and that all the studied models do not always work well. The problem of log-based anomaly detection has not been solved yet. Based on our findings, we also suggest possible future work.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Extensive machine learning and deep learning-based techniques have been proposed in research for automated log anomaly detection [9], [10], with the aim of reducing manual analysis costs.”
From Practitioners’ expectations on log anomaly detection (2025) · cites this paper as [10]
Experimental — accuracy not yet measuredThe check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“However, the adoption of LLM agents in the AIOps field, especially with noisy and lengthy data, remains limited [21,28].”
From RCAgent: Cloud Root Cause Analysis by Autonomous Agents with Tool-Augmented Large Language Models (2024) · cites this paper as [21]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 1×
“For example, Logsy [39] utilizes the Transformer network [46] to detect anomalies from log data.”
MethodExploiting Generative AI to Scale up Intelligent Tutoring Systems
Jakubův, J., Chvalovský, K., Goertzel, Z., Kaliszyk, C., Olšák, M., Piotrowski, B., Schulz, S., Suda, M., & Urban, J. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.itp.2023.19 · Cited by 76,658
Glove: Global Vectors for Word Representation
Pennington, J., Socher, R., & Manning, C. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1162 · Cited by 34,071 · cited 1×
“PLELog extracts the semantic vectors of log templates by using a pre-trained Glovemodel [41].”
MethodLearning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation
Cho, K., van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1179 · Cited by 25,545 · cited 1×
“Recurrent Neural Networks (RNNs), including its variants Long Short-Term Memory (LSTM) [21] and Gated Recurrent Units (GRUs) [9], are neural networks designed to handle the sequential inputs with arbitrary length.”
BackgroundConvolutional Neural Networks for Sentence Classification
Kim, Y. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1181 · Cited by 14,831 · cited 1×
“In [33], a CNN model originally designed for sentence classification [25] is applied to log-based anomaly detection.”
MethodLearning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
Cho, K., van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1406.1078 · Cited by 3,317 · cited 1×
“Recurrent Neural Networks (RNNs), including its variants Long Short-Term Memory (LSTM) [21] and Gated Recurrent Units (GRUs) [9], are neural networks designed to handle the sequential inputs with arbitrary length.”
Backgroundhdbscan: Hierarchical density based clustering
McInnes, L., Healy, J., & Astels, S. (2017) · The Journal of Open Source Software · doi:10.21105/joss.00205 · Cited by 2,833 · cited 2×
“PLELog adopts a clustering method (i.e., HDBSCAN [35]) to probabilistically estimate the labels of unlabeled log sequences.”
BackgroundBeyond Accuracy, F-Score and ROC: A Family of Discriminant Measures for Performance Evaluation
Sokolova, M., Japkowicz, N., & Szpakowicz, S. (2006) · Lecture Notes in Computer Science · doi:10.1007/11941439_114 · Cited by 1,974 · cited 1×
“Specificity [43, 49], which is the percentage of log sequences that are correctly identified normal over all real normal sequences, can measure the probability of false alarms.”
BackgroundDeepLog: Anomaly Detection and Diagnosis from System Logs through Deep Learning
Du, M., Li, F., Zheng, G., & Srikumar, V. (2017) · Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3133956.3134015 · Cited by 1,835 · cited 10×
“To answer the above questions, we conduct a systematic evaluation of five representative deep learning models for log-based anomaly detection (including DeepLog [12], LogAnomaly [36], PLELog [48], LogRobust [54], and CNN [33]) on four datasets (including HDFS, BGL, Spirit, and Thunderbird), under controlled experiment…”
BackgroundMethodInfluentialDetecting large-scale system problems by mining console logs
Xu, W., Huang, L., Fox, A., Patterson, D., & Jordan, M. I. (2009) · Proceedings of the ACM SIGOPS 22nd symposium on Operating systems principles · doi:10.1145/1629575.1629587 · Cited by 1,332 · cited 1×
“Over the years, many data-driven methods have been proposed to automatically detect system anomalies by analyzing log data [6, 7, 12, 29, 47, 50, 54].”
Research works printed but not linked: 2
- [21] Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural computation 9, 8 (1997), 1735-1780. -
- [31] XuLiu, Weiyou Liu, Xiaoqiang Di, Jinqing Li, Binbin Cai, Weiwu Ren, and Huamin Yang. 2021. LogNADS: Network anomaly detection scheme based on semantic representation. Future Generation Computer Systems (2021). -
Cited by
Items 101–200 of 260
UDAD: An Accurate Unsupervised Database Anomaly Detection Method
Zhong, H., Zhang, F., Zhao, Y., Zhang, W., Xiao, W., Tang, X., & Zang, L. (2023) · IEEE International Performance, Computing, and Communications Conference · doi:10.1109/ipccc59175.2023.10253824 · Cited by 4
Pluto: Sample Selection for Robust Anomaly Detection on Polluted Log Data
Ma, L., Cao, L., VanNostrand, P. M., Hofmann, D. M., Su, Y., & Rundensteiner, E. A. (2024) · Proceedings of the ACM on Management of Data · doi:10.1145/3677139 · Cited by 10
“In addition, these methods have been shown to degrade dramatically even if the training data is only slightly polluted, due to the corrupted learned representation [20] – which we confirm later in our experimental study (Sec.”
Transformer Models with Explainability for IT Telemetry and Business Events
Lim, S. H., & Wynter, L. (2024) · IEEE International Conference on Software Services Engineering (SSE) · doi:10.1109/sse62657.2024.00030 · Cited by 2
“We evaluate the TEVT model on two benchmark log anomaly datasets, BGL and Thunderbird, originally provided by [4] and further pre-processed by [5].”
BackgroundMethodInfluentialEDSLog: Efficient Log Anomaly Detection Method Based on Dataset Partitioning
Liang, F., & Liu, J. (2024) · Lecture notes in computer science · doi:10.1007/978-981-96-0602-3_22 · Cited by 2
System Log Anomaly Detection With Noise-Contrastive Learning and Pattern Feature
Luo, P., Deng, D., Xie, M., Yang, G., Chu, J., Soong, B., & Yuen, C. (2025) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2025.3579809 · Cited by 4
“These methods typically leverage Long Short-Term Memory (LSTM) networks [13], [15], [16], [17] to predict the next event in a log event sequence.”
GELog: a GPT-Enhanced Log Representation Method for Anomaly Detection
Xu, W., Wang, P., Shi, H., Zhou, G., Yao, J., & Zhang, X. (2025) · IEEE International Conference on Program Comprehension (ICPC) · doi:10.1109/icpc66645.2025.00062 · Cited by 3
“The implementations of DeepLog, LogAnomaly, and LogRobust were based on the publicly available source code by Zhang et al. [32] on GitHub.”
EarlyDLM: Early Detecting Lateral Movements Through Graph and Loopback Sequence Embedding
Hao, X., Liu, B., Wang, S., Cao, J., Zhou, D., & Fu, X. (2025) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2025.3584627 · Cited by 2
SSDALog: Semi-Supervised Domain Adaptation for Incremental Log-Based Anomaly Detection
Tian, J., Li, M., Chen, L., Wang, Z., Nie, X., & Qin, J. (2025) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2025.3583483 · Cited by 5
“2) Tool Comparison: In conducting our tool comparison, we have utilized a unified framework [18] built on PyTorch to implement static log anomaly detection methods, including LogAnomaly, LogRobust, and PLELog.”
InfluentialNipa, N. A., Bouguila, N., & Patterson, Z. (2025) · IEEE International Symposium on Industrial Electronics (ISIE) · doi:10.1109/isie62713.2025.11124715 · Cited by 3
“With advancements in deep learning methods [4], [5], [9] researchers have increasingly focused on enhancing both accuracy and efficiency of log-based anomaly detection methods [10].”
A Log Anomaly Detection Method Based on the Pre-Training and Fine-Tuning Framework
Li, G., Mo, J., Zhou, G., & Li, C. (2025) · IEEE China Conference on System Simulation Technology and its Applications (CCSSTA) · doi:10.1109/ieeeconf65522.2025.11136981 · Cited by 2
“Following previous studies [16] and our observations that HDFS is easier to deploy, we use HDFS as the source domain and select the more challenging BGL and OpenStack datasets as target domains.”
SemanticLog: Towards Effective and Efficient Large-Scale Semantic Log Parsing
Zhang, C., Xu, W., Liu, J., Zhang, L., Liu, G., Guan, J., Zhou, Q., & Zhou, S. (2025) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2025.3625121 · Cited by 7
“For example, Alibaba Inc.’s cloud computing systems generate approximately 30 to 50GB of logs per hour [37].”
Nasirzadeh, M., Tahmoresnezhad, J., & Khazaee, P. R. (2025) · Scientific Reports · doi:10.1038/s41598-025-27693-4 · Cited by 3
“Numerous research endeavors have examined log-based anomaly detection 1,20,22–24, 40 , 108 .”
DualBERT: Fusing Symbolic and Temporal Dynamics for High-Precision Log Anomaly Detection
Alzamil, A. A., Guesmi, R., AlGhadhban, A., Guesmi, T., & Kachouri, A. (2026) · IEEE Access · doi:10.1109/access.2026.3653996 · Cited by 2
K 4 : Online Log Anomaly Detection Via Unsupervised Typicality Learning
Chen, W., Singh, V., Rahmani, Z., Ganguly, D., Hariri, M., & Chaudhary, V. (2025) · IEEE International Conference on High Performance Computing, Data, and Analytics · doi:10.1109/hipc66333.2025.00019 · Cited by 12
Lightweight Techniques for Federated Anomaly Detection in Log Data
Pustozerova, A., Gómez, A. G., Landauer, M., Wurzenberger, M., Skopik, F., Weippl, E. R., Mayer, R. L., & Ekelhart, A. (2026) · IEEE Transactions on Reliability · doi:10.1109/tr.2026.3677674 · Cited by 4
Towards Massive Log Anomaly Detection Based on an Enhanced Multi-dimensional Time-domain LSTM
Wang, Y., Zhang, C., & Yu, J. (2023) · IEEE International Conference on Big Data Computing Service and Applications (BigDataService) · doi:10.1109/bigdataservice58306.2023.00042 · Cited by 3
LogDAPT: Log Data Anomaly Detection with Domain-Adaptive Pretraining (industry track)
Zheng, H., Chu, G., Sun, H., Wang, J., Tao, S., & Yang, H. (2023) · International Middleware Conference · doi:10.1145/3626562.3626830 · Cited by 4
“Although pretrained LMs proves to be effective in anomaly detection with abundant labeled log sequences, its performance could be severely undermined when there is not enough labeled data, which is frequently encountered in many industrial practices [11].”
BackgroundAutoLAD: An Unsupervised Anomaly Detection Method for Supercomputer Logs
Li, Y., Xiao, L., Feng, J., Zhang, J., Ma, Q., & Yuan, Y. (2023) · International Conference on Electronics, Computers and Communication Technology (CECCT) · doi:10.1145/3637494.3637501 · Cited by 2
Optimizing Software Release Management with GPT-Enabled Log Anomaly Detection
Mannam, P. K. (2023) · International Conference on Computer and Information Technology (ICCIT) · doi:10.1109/iccit60459.2023.10441522 · Cited by 3
Cloud RAN Based Privacy Preserving Federated Cross Domain Anomaly Detection in IoT Devices Logs
Bashir, Z., Kumar, S., Murgai, V., & Singh, S. (2023) · IEEE Globecom Workshops (GC Wkshps) · doi:10.1109/gcwkshps58843.2023.10464603 · Cited by 2
“We use the HDFS dataset for the central server, Hadoop, BGL, Spirit, Thunderbird [5] for client 1, client 2, client 3 and client 4 respectively.”
MethodLog Anomaly Detection in Application Servers Using Deep Learning
Alagöz, E., Şahin, Y. M., Karaca, K. Ö., Gedik, A. O., & Akcayol, M. A. (2024) · Engineering cyber-physical systems and critical infrastructures · doi:10.1007/978-3-031-56310-2_20 · Cited by 3
Landscape and Taxonomy of Online Parser-Supported Log Anomaly Detection Methods
Lupton, S., Washizaki, H., Yoshioka, N., & Fukazawa, Y. (2024) · IEEE Access · doi:10.1109/access.2024.3387287 · Cited by 9
“Le and Zhang performed numerous experiments using deep learning log anomaly detection methods to analyze the impact of training data strategies, grouping approaches, class distributions, and data noise [247].”
BackgroundMachine Learning and Anomaly Detection for an Automated Monitoring of Log Data
Falzone, S., Gühring, G., & Jung, B. (2024) · Springer eBooks · doi:10.1007/978-3-031-57452-8_13 · Cited by 3
Unveiling Hidden Anomalies: Leveraging SMAC-LSTM for Enhanced Software Log Analysis
Sun, Y., Keung, J., Zhang, J., Yu, H. K., Luo, W., & Liu, S. (2024) · IEEE Annual International Computer Software and Applications Conference (COMPSAC) · doi:10.1109/compsac61105.2024.00157 · Cited by 6
“Current deep learning-based anomaly detection models typically convert logs into three primary types [1]: (1) Sequential vectors, (2) Quantitative vectors, and (3) Semantic vectors.”
BackgroundLu, R., Zhu, X., Li, X., Long, N., & Zhang, G. (2024) · International Conference on Machine Learning and Computer Application (ICMLCA) · doi:10.1109/icmlca63499.2024.10753853 · Cited by 2
“By analysing log data, it can help locate system anomalies, performance bottlenecks, and potential faults [6].”
Effective IDS under constraints of modern enterprise networks: revisiting the OpTC dataset
Nikulshin, V., & Talhi, C. (2024) · Conference on Cloud and Internet of Things (CIoT) · doi:10.1109/ciot63799.2024.10757066 · Cited by 3
“Each step contributes to the loss of information, and the devastating effects of label errors have been analysed at length, for example in [18], [19].”
SeLog: A Log Anomaly Detection Method on Log Event and Variable Semantic
Li, C., Chen, G., Mi, X., Zhang, Y. E., & Zhou, H. (2024) · IEEE International Conference on Information Systems and Computer Aided Education (ICISCAE) · doi:10.1109/iciscae62304.2024.10761356 · Cited by 2
“Therefore, fully mining log information to implement efficient anomaly detection has become an urgent need [3].”
Towards Hybrid Architectures for Big Data Analytics: Insights From Spark-MPI Integration
Zhou, M., Li, Q., Cai, M., Xu, C., & Wang, Y. (2025) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2025.3562342 · Cited by 2
“Examples include large-scale data analytics, such as log analysis [30], ETL (Extract, Transform, Load) processes [31], and data cleansing [32].”
Large Language Models Empowered Online Log Anomaly Detection in AIOps
Zhang, S., Fan, D., & He, L. (2024) · Asia-Pacific Software Engineering Conference (APSEC) · doi:10.1109/apsec65559.2024.00051 · Cited by 5
LogLSHD: Fast Log Parsing with Locality-Sensitive Hashing and Dynamic Time Warping
Huang, S.-W., Wu, X., & Li, H. (2025) · doi:10.1145/3727582.3728685 · Cited by 3
United We Stand: Towards End-to-End Log-based Fault Diagnosis via Interactive Multi-Task Learning
He, M., Duan, C., Xiao, P., Jia, T., Yu, S., Zhang, L., Hong, W., Han, J., Wu, Y., Li, Y., et al. (2025) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase63991.2025.00061 · Cited by 11
Verification of Generalizability in Software Log Anomaly Detection Models
Uchida, H., Tominaga, K., Itai, H., Li, Y., & Nakatoh, Y. (2024) · Artificial intelligence · doi:10.5772/intechopen.111938 · Cited by 2
Fiber Optical Module Anomaly Detection Using Graph Deep Learning Model
Li, Y., Li, J.-Y., Kao, H.-Y., & Tsai, Y.-L. (2024) · International Conference on Computer, Information and Telecommunication Systems · doi:10.1109/cits61189.2024.10607993 · Cited by 1
Event-level Anomaly Detection on Software logs: Role of Algorithm, Threshold, and Window Size
Nyyssölä, J., & Mäntylä, M. (2024) · IEEE International Conference on Software Quality, Reliability and Security (QRS) · doi:10.1109/qrs62785.2024.00070 · Cited by 1
“Additionally, the models are faster to train [10].”
BackgroundPre-LogMGAE: Identification of Log Anomalies Using a Pre-Trained Masked Graph Autoencoder
Wu, A., & Kwon, Y. (2024) · IEEE Symposium on Reliable Distributed Systems (SRDS) · doi:10.1109/srds64841.2024.00036 · Cited by 1
Wang, P., Zhang, X., & Cao, Z. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-024-06850-y · Cited by 2
Not All Exceptions Are Created Equal: Triaging Error Logs in Real-World Enterprises
Liu, J., Yao, M., Li, S., Yang, D., Wu, Z., Qu, X., Zhang, Z., Li, D., Guo, Y., & Chen, X. (2025) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3721126 · Cited by 4
LogSI: A Benchmark for System-Incremental Log Analysis
Zhou, M., Yang, W., Ma, L., Jiang, S., Xu, B., & Xiao, Y. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890641 · Cited by 1
“For log anomaly detection, we use the dataset and settings from LogADEmpirical [34], which includes logs from 4 systems.”
LogProb: Online Parsing Evolving Logs With Complex Parameters
Li, Y., Wang, Z., Mao, J., Gu, Z., & Jiang, H. (2025) · IEEE Access · doi:10.1109/access.2025.3569548 · Cited by 1
“For example, Alibaba Inc.’s cloud computing system generates about 30-50 gigabytes (about 120 to 200 million lines) of trace logs per hour [4].”
Wu, X., Li, H., & Khomh, F. (2025) · Automated Software Engineering · doi:10.1007/s10515-025-00527-3 · Cited by 5
ALMOND: Learning an Assembly Language Model for 0-Shot Code Obfuscation Detection
Li, X., Yu, S., & Yin, H. (2025) · Proceedings of the ACM on software engineering. · doi:10.1145/3728886 · Cited by 2
“However, we have already observed that combining perplexity and error-perplexity outperforms using either metric alone.”
Anomaly Detection in Large-Scale Monitoring Systems using a Language Model
Vorathammathorn, S., Subsa-Ard, N., Thaepprasit, T., Phunchongharn, P., & Tarnpradab, S. (2025) · doi:10.1145/3718350.3718354 · Cited by 2
“Addressing this limitation, deep learning approaches have gained in importance, convincing results in software log analysis [2, 3, 5, 6, 8, 9, 12, 15, 18].”
Aljumaily, M. S., & Abd, H. (2025) · International Journal of Mechatronics Robotics and Artificial Intelligence · doi:10.33971/ijmrai.1.2.11 · Cited by 4
FedLAD: A Modular and Adaptive Testbed for Federated Log Anomaly Detection
Liao, Y., Keung, J., Mao, Z., Zhang, J., & Li, J. (2025) · IEEE International Conference on Autonomic Computing and Self-Organizing Systems Companion (ACSOS-C) · doi:10.1109/acsos-c66519.2025.00066 · Cited by 1
“As a result, L og-based A nomaly D etection (LAD) has become a vital approach for automatically identifying abnormal behaviors and potential faults [2], and it has been successfully applied in diverse domains such as data centers, Internet of Things, and autonomous vehicles [3].”
Too Many Cooks: Assessing the Need for Multi-Source Data in Microservice Failure Diagnosis
Zhang, S., Feng, X., Wang, R., Ma, M., Gu, W., Sun, Y., Jia, Z., Sun, J., & Pei, D. (2025) · Proceedings/Proceedings - International Symposium on Software Reliability Engineering · doi:10.1109/issre66568.2025.00014 · Cited by 5
“In recent years, there have been numerous empirical studies focusing on failure management based on a single source of data [47], [48], [49], [50], [51], [52].”
GLU-PM: A Pretrained Model Toward Generalized Log Understanding
Liu, X., Zheng, S., Zou, H., Yu, H., & Gao, S. (2025) · IEEE Transactions on Reliability · doi:10.1109/tr.2025.3642425 · Cited by 2
GNN-driven dynamic access control for context-embedded neighborhood fusion
Chen, H., Yuan, L., Bao, H., Dai, H., Xiang, Y., & Wang, K. (2025) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-025-00399-3 · Cited by 1
“Le and Zhang (2022) propose DeepLog model to capture log sequence anomalies using LSTM, Chen et al. (2021) use convolutional neural networks to automatically mine log features.”
RELog: Robust and Efficient Anomaly Detection Based on Complete Logs for Large-Scale Systems
Chai, X., Feng, X., Lou, Z., Sun, Y., & Guizani, M. (2026) · IEEE Transactions on Computers · doi:10.1109/tc.2026.3654206 · Cited by 2
“However, research [18] suggests that supervised approaches incorporating abnormal logs often outperform purely semi-supervised ones.”
Root Cause Analysis of RISC-V Build Failures via LLM and MCTS Reasoning
Shuai, W., Liu, J., Ma, Z., Kang, L., Wang, Z., Wang, S., Ye, D. Q., Li, H., Wang, W., & Zhu, J. (2025) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase63991.2025.00227 · Cited by 1
“Log Anomaly Detection (LAD) is a critical task in information technology [8], [9] that aims to identify anomalous log entries or sequences that deviate from normal patterns within massive log data to uncover potential system issues or abnormal behaviors.”
LogMoE: Lightweight Expert Mixture for Cross-System Log Anomaly Detection
Qi, J., Luan, Z., Huang, S., Fung, C., Wang, Y., Wang, A., Zhang, H., Yang, H., & Qian, D. (2025) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase63991.2025.00035 · Cited by 3
Self-citationFuseLog: Fusing Symbolic and Temporal Dynamics for Enhanced Log Anomaly Detection
Alzamil, A. A., Guesmi, R., Gasmi, T., Marouani, I., AlGhadhban, A., Alshammari, B. M., Guesmi, T., Fadhlaoui, M. K., & Guermazi, N. (2026) · IEEE Access · doi:10.1109/access.2026.3665405 · Cited by 1
GPT-based self-supervised anomaly detection in command lines
Li, M. Q., Keutchayan, J., Charest, F., & Fung, B. C. M. (2026) · Journal of Computer Virology and Hacking Techniques · doi:10.1007/s11416-026-00606-3 · Cited by 1
A Stage-Aware Requirements-To-Evidence Framework for Secure CI/CD Pipelines
Saleh, S. M., Madhavji, N. H., & Steinbacher, J. (2026) · IEEE International Requirements Engineering Conference Workshops (REW) · doi:10.1109/rew72749.2026.00036 · Cited by 1
“Le and Zhang’s empirical assessment cautions that deep-learning-based log anomaly detection has practical limitations, which supports the framework’s decision to treat anomaly output as triage evidence rather than final requirements-level proof [20].”
Optimizing Anomaly Detection in Large-scale Logs
Techaviseschai, A., Barroso, V. C., Tarnpradab, S., & Phunchongharn, P. (2023) · International Conference on Information and Communication Technology Convergence (ICTC) · doi:10.1109/ictc58733.2023.10392401 · Cited by 1
“Apart from CNN applications, RNN and its variation such as LSTM have also been used in several works [13, 14].”
MethodLog Anomaly Detection Based on Semantic Features and Topic Features
Wang, P., Zhang, X., & Cao, Z. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-0808-6_24 · Cited by 1
Multi-factor Acquisition of Global State based Log-anomaly Detection
Lu, Z., & Zeng, X. (2023) · International Conference on Electronic Information Technology and Computer Engineering (EITCE) · doi:10.1145/3650400.3650608 · Cited by 1
Engineering Industry-Ready Anomaly Detection Algorithms
Nguyen, N.-T., Heldal, R., & Pelliccione, P. (2024) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3639478.3643085 · Cited by 1
LogTIW:A log anomaly detection model based on TF-IDF weighted semantic features
Kang, J., Zhao, J., & Li, Z. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10651316 · Cited by 1
Optimizing Software Release Management with GPT-Enabled Log Anomaly Detection
Setu, J. H., Hossain, M. S., Halder, N., Islam, A., & Amin, M. A. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-78166-7_23 · Cited by 3
VCRLog: Variable Contents Relationship Perception for Log-based Anomaly Detection
Wang, J., Li, T., Zhang, R., Tang, Z., Wu, D., & Yang, Z. (2024) · Proceedings/Proceedings - International Symposium on Software Reliability Engineering · doi:10.1109/issre62328.2024.00025 · Cited by 3
Research on Fault Prediction of Computer Network Nodes Based on Log information
Wang, Y., Ye, Q., Zhou, P., Li, B., Wang, X., & Zhang, L. (2024) · IEEE International Conference on Communication Software and Networks · doi:10.1109/iccsn63464.2024.10793322 · Cited by 1
Early Detection of Malicious Activity in Log Event Sequences Using Deep Learning
Ralbovský, A., & Kotuliak, I. (2025) · Conference of Open Innovations Association (FRUCT) · doi:10.23919/fruct65909.2025.11008245 · Cited by 1
E-Log: Fine-Grained Elastic Log-Based Anomaly Detection and Diagnosis for Databases
Zhang, L., Jia, T., Tan, X., Huang, X., Jia, M., Liu, H., Wu, Z., & Li, Y. (2025) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2025.3594870 · Cited by 12
“Log grouping organizes logs into sequence groups using fixed time windows [30].”
Zhang, H., Yu, D., Zhang, L., Rong, G., Yu, Y., Shen, H., Zhang, H., Shao, D., & Kuang, H. (2025) · Asia-Pacific Symposium on Internetware · doi:10.1145/3755881.3755889 · Cited by 7
Self-citationToward Efficient Anomaly Detection on Edge Devices with KnowledgePreprint
Nguyen, H., Nguyen, L.-V., & Nguyen, D.-H. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4526761
SSADLog: Whole Lifecycle Tuning Anomaly Detection with Small Sample LogsPreprint
Zhou, Z., Zhou, M., Wu, A., Xia, J., & Liu, W. (2023) · Research Square · doi:10.21203/rs.3.rs-3588406/v1
Self-supervised Federated Learning for Anomaly Detection
Eng, T.-Q., Pao, H.-K., & Liao, C.-C. (2023) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata59044.2023.10386871 · Cited by 1
“Regarding anomaly detection, former studies [2], [3], have demonstrated that system log data exhibit a highly skewed distribution, with anomalies representing a small portion of the dataset.”
BackgroundDevice Performance Anomaly Detection Method Based on Graph Convolutional Neural Network
Liu, A., Yang, Y., Guo, Y., Gao, Z., & Rui, L. (2024) · Lecture notes in electrical engineering · doi:10.1007/978-981-99-9243-0_24
Automated Log-Based Anomaly Detection under Noise Circumstances
Zhu, Y., Min, G., Wu, Y., & Wang, H. (2023) · IEEE Smart World Congress (SWC) · doi:10.1109/swc57546.2023.10448977
Non-Deterministic and Semi-Supervised Log-based Anomaly Detection
Kumar, S., Murgai, V., Bashir, Z., & Singh, S. (2023) · IEEE Globecom Workshops (GC Wkshps) · doi:10.1109/gcwkshps58843.2023.10464781
“2) Semantic Embedding: Log statements are not static and can change over time to capture new information and improve the overall quality of system logs [8] [3] [1] [12].”
BackgroundMethodInfluentialAutomatic Anomaly Detection Using Unlabelled Log Data
Zhu, Y., Min, G., Wu, Y., & Wang, H. (2023) · IEEE International Conference on High Performance Computing and Communications; IEEE International Conference on Smart City; IEEE International Conference on Data Science and Systems (HPCC/SmartCity/DSS) · doi:10.1109/hpcc-dss-smartcity-dependsys60770.2023.00075
LogFold: Enhancing Log Anomaly Detection Through Sequence Folding and Reconstruction
Shi, X., Li, R., Du, Q., He, C., & Tian, F. (2023) · Asia-Pacific Software Engineering Conference (APSEC) · doi:10.1109/apsec60848.2023.00028 · Cited by 1
“Following the typical deep learning log anomaly detection framework [11], LogFold consists of four steps: (1) Log Parsing and Grouping .”
MethodSSADLog: Whole Lifecycle Tuning Anomaly Detection with Small Sample LogsPreprint
Zhou, Z., Zhou, M., Wu, A., Xia, J., & Liu, W. (2024) · Research Square · doi:10.21203/rs.3.rs-3588406/v2
Towards an Approach for Representing Log Event in Anomaly Detection
Nguyen, H., Le, H.-V., Minh, B. K., Huan, P. V., & Tan, N. H. A. (2024) · International Conference on Expert Clouds and Applications (ICOECA) · doi:10.1109/icoeca62351.2024.00030
“However, with different developers and different system scales and sizes, a large amount of log formats are used and generated during its operation [2].”
BackgroundLog Parsing using Semantic Filtering based Prompt Learning
Xu, Z., Huang, B., & Chen, N. (2024) · IEEE International Conference on Software Quality, Reliability and Security (QRS) · doi:10.1109/qrs62785.2024.00072 · Cited by 1
“Many researchers have made significant efforts in the field of log based fault diagnosis [2].”
BackgroundDesign Graph-Structured Dataset and Feature Selection for Cyber Threat Detection
Boonyopakorn, P., & Changsan, U. (2024) · Lecture notes in networks and systems · doi:10.1007/978-3-031-73122-8_12
Research on Anomaly Detection Based on System Log Data
Zhang, B., & Jiang, Y. (2024) · International Conference on Computer Science and Management Technology (ICCSMT) · doi:10.1145/3708036.3708050
Hierarchical Prompt Tuning for System-Incremental Log Analysis
Zhou, M., Yang, W., Ma, L., Jiang, S., Xu, B., & Xiao, Y. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890029
ARM Server Error Prediction based on Fusion Features
Zhang, C., Ji, S., Zhang, M., & Jiang, Z. (2024) · IEEE International Conference on Frontiers Technology of Information and Computer (ICFTIC) · doi:10.1109/icftic64248.2024.10912930
“Log prediction methods can identify anomalous sequences but are deficient in distinguishing different types of errors, primarily due to the complexity and diversity of log data [8-9] .”
Log-Based Representation Transferable Learning for Cross-System Anomaly Detection
Nguyen, M.-T., Dinh, T.-D. L., & Cao, V. L. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-4285-4_27
When 5G Security Meets Natural Language Processing: Challenges and Future Pathways
Nguyen, H., Nguyen, N. H., Dao, N. K., & Bao, T. Q. (2024) · IEEE Conference on Antenna Measurements and Applications (CAMA) · doi:10.1109/cama62287.2024.10986050
“The authors in [6] creatively combined the use of NLP models with network packet log files, which resemble a direction of research that has only been recently delved into [26].”
Unsupervised Log Parsing Based on Large Language Models and Entropy
Duan, Y., Xu, J., Fan, C., & Liu, Z. (2025) · International Symposium on System Security, Safety, and Reliability (ISSSR) · doi:10.1109/isssr65654.2025.00013
MADM: Microservice Anomaly Detection Based on Multi-source Data
Li, Y., & Ding, Y. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-6966-0_27
LogMT: A Self-supervised Log Anomaly Detection Method Based on Multi-Tasks
Xie, W., & Ni, Q. (2025) · International Conference on Computer Supported Cooperative Work in Design (CSCWD) · doi:10.1109/cscwd64889.2025.11033424
“In this context, high availability and reliability are essential for large-scale software-intensive systems [1].”
Automated, Unsupervised, and Auto-Parameterized Inference of Data Patterns and Anomaly Detection
Qin, Q., Li, H., Merlo, E., & Lamothe, M. (2025) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse55347.2025.00078 · Cited by 2
Uchida, H., Tominaga, K., Itai, H., Li, Y., & Nakatoh, Y. (2025) · Pertanika journal of science & technology · doi:10.47836/pjst.33.s4.09 · Cited by 1
Bandlamudi, A. K., & Pachala, S. (2025) · Lecture notes in networks and systems · doi:10.1007/978-981-96-6063-6_10 · Cited by 1
LogLabeler: Towards Effective Acquisition of Log Labels in Industrial Log-Based Analysis
Li, Z., Wang, Q., Cai, S., Liu, Z., Ma, T., Yang, W., Li, Y., & Xie, T. (2025) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2025.3605635
“In the literature, A is often binary for anomaly detection [11], [12]; however, industry may use finer-grained levels (e.g., 0-3 for normal to severe anomalies).”
InfluentialSemi-supervised Log Anomaly Detection Method Based on Uncertainty-Aware Self-training
Wang, R., Zheng, R., Li, Y., Yan, L., Shen, J., & Liu, J. (2025) · Lecture notes in networks and systems · doi:10.1007/978-3-031-94892-3_11
Log Anomaly detection in IoT networks using DeBERTa-V3
Haralambie, F., Bratu, A., Raluca-Ioana, P., Nicolae, G., & Vasilăţeanu, A. (2025) · IEEE International Conference on Blockchain, Smart Healthcare and Emerging Technologies (SmartBlock4Health) · doi:10.1109/smartblock4health64843.2025.11189737
A Novel Semi-Supervised Model for Generalizing Log Anomaly Detection with Limited Labeled Data
Sun, Y., Keung, J., Yang, Z., Liu, S., & Chan, W. K. (2025) · IEEE International Conference on Software Quality, Reliability and Security Companion (QRS-C) · doi:10.1109/qrs-c65679.2025.00103
“Unlike other models, which often struggle with high variability in log formats and system behaviors [4], GANs do not rely on predefined rules or fixed patterns, making them more adaptable to the inherent diversity of log data.”
A Data-Driven Framework for Anomaly Detection in Industrial Systems Using Log Data
Hewing, M., Zhao, Y., Vossel, M., Nieschler, P., & Kleinert, T. (2025) · Annual Conference of the IEEE Industrial Electronics Society (IECON) · doi:10.1109/iecon58223.2025.11221513
“To address the limitations of classical machine learning approaches, such as inefficiency and weak adaptability, research has increasingly focused on deep learning for log-based anomaly detection [16].”
LogReader: General-Purpose Log Analysis via Open-Source Large Language Models
Qi, J., Luan, Z., Huang, S., Lu, Y. H., Yang, H., & Qian, D. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11227375 · Cited by 1
Anomaly Detection With Machine Learning
Shavali, T. K. S., Mahale, S. V., Manikandan, S., Chaudhari, S. V., Kunwar, F. B., Sandeep, C. S., & Bhoopathy, V. (2025) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-2252-0.ch004
FLUXLog: A Federated Mixture-of-Experts Framework for Unified Log Anomaly Detection
Xia, Y., Zhao, Y., Wan, J., & Jiang, C. (2025) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2025.3638693 · Cited by 1
LogMerge: improved log parsing based on two-step clustering combined with low-level token processing
Le, H.-V., Nguyen, H., Huynh, N. K., & Phan, H. D. (2025) · International Journal of Information Technology · doi:10.1007/s41870-025-02911-4
Enhancing Fault Detection in Mechatronic Systems Utilizing the Transition Analysis Algorithm
Hoffmann, F., Heine, J., & Bäker, B. (2025) · International Conference on Electrical Engineering and Mechatronics Technology (ICEEMT) · doi:10.1109/iceemt66565.2025.11280700
“…method for anomaly detection in log data meets the current standard for statistical methods and can likely compete with deep learning approaches [3-7 , After investigating the use of transition analysis for anomaly detection in raw sensor data in [10], it has now been shown to be applicable for…”
Anomaly Detection in Log Data: A Comparative Study
Sedláček, O., Žádník, M., & Bartoš, V. (2025) · International Conference on Network and Service Management (CNSM) · doi:10.23919/cnsm67658.2025.11297503 · Cited by 1
“For example, NeuralLog applies sliding windows of 20 lines without component boundaries [16]; Guo et al. use a 5-minute time window followed by 128-line splits [7]; and Le et al. compare fixed windows of 20, 100, and 200 lines, concluding that component grouping performs best [31].”
DLogParser: An Efficient Dynamic Log Parser with Multiple Grouping Criteria
Yuan, J., Wang, C., Zhou, H., Zhang, Y., & Wang, Y. (2026) · Applied Sciences · doi:10.3390/app16020811
Metadata
- Authors
- Le, V.-H., & Zhang, H.
- Affiliations
- University of Newcastle Australia (AU)
- Venue
- Proceedings of the 44th International Conference on Software Engineering
- Publisher
- ACM
- Published
- 2022-05-21
- Pages
- 1356-1367
- ISSN
- 0270-5257 (ISSN-L)
- OpenAlex
- W4226128225
- Cited by
- 257
- References
- 51
- Citations by year
- 2026: 59 · 2025: 88 · 2024: 57 · 2023: 49 · 2022: 4
- FWCI
- 47.97
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Funders
- Australian Research Council?s Discovery Projects (DP200102940)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › Software System Performance and Reliability
- Physical Sciences › Computer Science › Artificial Intelligence › Anomaly Detection Techniques and Applications
- Physical Sciences › Computer Science › Software › Software Testing and Debugging Techniques
- Keywords
- log anomaly detection, deep learning models, model evaluation, training data selection, data grouping, class distribution, noisy data, early detection
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.