Paper: arXiv 2403.00775

Abstract

Detecting anomalies is important for identifying inefficiencies, errors, or fraud in business processes. Traditional process mining approaches focus on analyzing ‘flattened’, sequential, event logs based on a single case notion. However, many real-world process executions exhibit a graph-like structure, where events can be associated with multiple cases. Flattening event logs requires selecting a single case identifier which creates a gap with the real event data and artificially introduces anomalies in the event logs. Object-centric process mining avoids these limitations by allowing events to be related to different cases. This study proposes a novel framework for anomaly detection in business processes that exploits graph neural networks and the enhanced information offered by object-centric process mining. We first reconstruct and represent the process dependencies of the object-centric event logs as attributed graphs and then employ a graph convolutional autoencoder architecture to detect anomalous events. Our results show that our approach provides promising performance in detecting anomalies at the activity type and attributes level, although it struggles to detect anomalies in the temporal order of events.

Complexity vs Empirical Score

  • Math Complexity: 7.0/10
  • Empirical Rigor: 8.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced mathematics, including graph theory, graph convolutional autoencoders, and formal definitions for object-centric event logs, indicating high math complexity. It demonstrates strong empirical rigor by evaluating on two public datasets (one synthetic, one real-life), injecting anomalies, and using standard metrics, with code and data available on GitHub.

Research Flowchart

  flowchart TD
  A["Research Goal: Detect anomalies<br>in object-centric<br>business processes"] --> B["Data Input:<br>Object-centric Event Logs"]
  B --> C["Methodology:<br>Reconstruct Process<br>Dependencies as<br>Attributed Graphs"]
  C --> D["Computation:<br>Graph Convolutional<br>Autoencoder"]
  D --> E["Outcomes: Detect anomalies in<br>Activity Types & Attributes"]
  D --> F["Limitation: Struggles with<br>Temporal Order Anomalies"]

Cited References (99)

  1. Semi- Supervised Classification with Graph Convolutional Networks (2016) — link
  2. Anomaly detection in high-energy physics using a quantum autoencoder (2022) — link
  3. Model- Agnostic Interpretability of Machine Learning (2016) — link
  4. Financial process mining - Accounting data structure dependent control flow inference (2017) — link
  5. A business process mining application for internal transaction fraud mitigation (2011) — link
  6. Internal fraud risk reduction: Results of a data mining case study (2010) — link
  7. Data mining journal entries for fraud detection: An exploratory study (2010) — link
  8. Cluster Analysis for Anomaly Detection in Accounting Data : An Audit Approach (2011) — link
and 91 more
  1. Detection of Accounting Anomalies in the Latent Space using Adversarial Autoencoder Neural Networks (2019) — link
  2. Detection of Anomalies in Large Scale Accounting Data using Deep Autoencoder Networks (2018) — link
  3. Defining Cases and Variants for Object - Centric Event Data (2022) — link
  4. ocpa: A Python library for object-centric process analysis (2022) — link
  5. Precision and Fitness in Object - Centric Process Mining (2021) — link
  6. Addressing Convergence , Divergence , and Deficiency Issues (2023)
  7. Deep learning (2016)
  8. Process Mining Manifesto (2012) — link
  9. Discovering Object -centric Petri Nets (2020) — link
  10. Conformance checking of processes based on monitoring real behavior (2008) — link
  11. Building instance graphs for highly variable processes (2016) — link
  12. Anomaly Detection Using Process Mining (2009) — link
  13. Graph- Based Feature Extraction on Object - Centric Event Logs (2022) — link
  14. Process Mining : Data Science in Action (2016) — link
  15. Login. MicroStrategy — link
  16. On the Bottleneck of Graph Neural Networks and its Practical Implications (2020) — link
  17. On the Bottleneck of Graph Neural Networks and its Practical Implications (2021) — link
  18. Exploring Gated Graph Sequence Neural Networks for Predicting Next Process Activities (2022) — link
  19. Analysis of Language Inspired Trace Representation for Anomaly Detection (2020) — link
  20. Aligning Data - Aware Declarative Process Models and Event Logs (2021) — link
  21. Anomaly Detection on Event Logs with a Scarcity of Labels (2020) — link
  22. Anomaly Detection in Business Process Runtime Behavior – Challenges and Limitations (2017) — link
  23. Demystifying Noise and Outliers in Event Logs : Review and Future Directions (2022) — link
  24. Anomaly Detection in Business Process Runtime Behavior – Challenges and Limitations (2017) — link
  25. How to Write an Abstract
  26. Editor’s Comments : Why Top Journals Accept Your Paper (2009) — link
  27. This is Not an Article Just Some Thoughts on How to Write One
  28. How to Write Beautiful Process -and- Data - Science Papers ? (2022) — link
  29. An End -to- End Process Model for Supervised Machine Learning Classification : From Problem to Deployment in Information Systems
  30. Introduction to Design Science Research (2020) — link
  31. Positioning and Presenting Design Science Research for Maximum Impact (2013) — link
  32. DutchNetSci Info Page — link
  33. LSTM - Based Anomaly Detection of Process Instances : Benchmark and Tweaks (2023) — link
  34. GRASPED : A GRU - AE Network Based Multi - Perspective Business Process Anomaly Detection Model (2023) — link
  35. Keeping our rivers clean: Information -theoretic online anomaly detection for streaming business process events (2022) — link
  36. Algorithms for anomaly detection of traces in logs of process aware information systems (2013) — link
  37. A Systematic Review of Anomaly Detection for Business Process Event Logs (2023) — link
  38. Autoencoders for improving quality of process event logs (2019) — link
  39. Aligning Data - Aware Declarative Process Models and Event Logs (2021) — link
  40. Analyzing business process anomalies using autoencoders (2018) — link
  41. BINet : Multi -perspective business process anomaly classification (2022) — link
  42. Object-centric behavioral constraint models: a hybrid model for behavioral and data perspectives (2019) — link
  43. OC - PM : analyzing object-centric event logs and process models (2022) — link
  44. BPI Challenge 2017 (2017) — link
  45. Object- Centric Process Mining : Dealing with Divergence and Convergence in Event Data (2019) — link
  46. OCEL : A Standard for Object - Centric Event Logs (2021) — link
  47. A Framework for Extracting and Encoding Features from Object - Centric Event Data (2022) — link
  48. Object-centric process predictive analytics (2023) — link
  49. Process Mining : Discovery , Conformance and Enhancement of Business Processes (2011) — link
  50. Process mining: A 360 degree overview (2022) — link
  51. An Event Data Extraction Approach from SAP ERP for Process Mining (2022) — link
  52. ACFE Report to the Nations : 2022 Global Fraud Study (2022) — link
  53. Graph Neural Networks : Foundations , Frontiers , and Applications (2022) — link
  54. GraphSVX : Shapley Value Explanations for Graph Neural Networks (2021) — link
  55. Higher-order Structure Based Anomaly Detection on Attributed Networks (2021) — link
  56. Recommending Related Products Using Graph Neural Networks in Directed Graphs (2023) — link
  57. Geometric Deep Learning : Grids , Groups , Graphs , Geodesics , and Gauges (2021) — link
  58. A Deep Learning Approach to Antibiotic Discovery (2020) — link
  59. Anomaly Detection using Graph Neural Networks (2019) — link
  60. Graph Convolutional Neural Networks for Web - Scale Recommender Systems (2018) — link
  61. Graph Neural Networks for Social Recommendation (2019) — link
  62. Laplacian Eigenmaps for Dimensionality Reduction and Data Representation (2003) — link
  63. Symmetric Graph Convolutional Autoencoder for Unsupervised Graph Representation Learning (2019) — link
  64. Graph Neural Networks for temporal graphs: State of the art, open challenges, and opportunities (2023) — link
  65. Benchmarking Graph Neural Networks
  66. Everything is Connected : Graph Neural Networks (2023) — link
  67. Graph Anomaly Detection with Graph Neural Networks : Current Status and Challenges (2022) — link
  68. Semi- Supervised Classification with Graph Convolutional Networks (2017) — link
  69. Variational Graph Auto - Encoders (2016) — link
  70. The Graph Neural Network Model (2009) — link
  71. Graph Autoencoders for Business Process Anomaly Detection (2021) — link
  72. Process Discovery Using Graph Neural Networks (2021) — link
  73. Exploiting Instance Graphs and Graph Neural Networks for Next Activity Prediction (2022) — link
  74. A Comparison of Deep - Learning Methods for Analysing and Predicting Business Processes (2021) — link
  75. A technique for determining relevance scores of process activities using graph-based neural networks (2021) — link
  76. Explainable predictive business process monitoring using gated graph neural networks (2020) — link
  77. Estimating the Contamination Factor ’s Distribution in Unsupervised Anomaly Detection (2022) — link
  78. A new non-parametric detector of univariate outliers for distributions with unbounded support (2017) — link
  79. PyOD : A Python Toolbox for Scalable Outlier Detection
  80. The relationship between Precision - Recall and ROC curves (2006) — link
  81. Outlier detection: how to threshold outlier scores? (2019) — link
  82. Statistical Independence and Novelty Detection with Information Preserving Nonlinear Maps (1996) — link
  83. A Causal Approach to Detecting Multivariate Time -series Anomalies and Root Causes (2023) — link
  84. Anomaly Detection Using Autoencoders with Nonlinear Dimensionality Reduction (2014) — link
  85. Conditional Anomaly Detection (2007) — link
  86. Isolation Forest (2008) — link
  87. Support Vector Method for Novelty Detection (1999) — link
  88. BOND : Benchmarking Unsupervised Outlier Node Detection on Static Attributed Graphs (2022) — link
  89. Anomaly detection: A survey (2009) — link
  90. Outlier Analysis (2013) — link
  91. An Introduction to Variational Autoencoders (2019) — link

0 of 99 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.