Paper: arXiv 2402.09495

Authors: Catayoun Azarm, Erman Acar, Mickey van Zeelt

Abstract

Online transaction fraud presents substantial challenges to businesses and consumers, risking significant financial losses. Conventional rule-based systems struggle to keep pace with evolving fraud tactics, leading to high false positive rates and missed detections. Machine learning techniques offer a promising solution by leveraging historical data to identify fraudulent patterns. This article explores using the personalised PageRank (PPR) algorithm to capture the social dynamics of fraud by analysing relationships between financial accounts. The primary objective is to compare the performance of traditional features with the addition of PPR in fraud detection models. Results indicate that integrating PPR enhances the model’s predictive power, surpassing the baseline model. Additionally, the PPR feature provides unique and valuable information, evidenced by its high feature importance score. Feature stability analysis confirms consistent feature distributions across training and test datasets.

Complexity vs Empirical Score

  • Math Complexity: 4.5/10
  • Empirical Rigor: 6.5/10
  • Quadrant: Street Traders — practical and empirical, lighter on theory

Why this score: The paper applies Personalised PageRank (PPR) to model transaction networks for fraud detection, providing practical empirical validation and stability analysis. While the graph algorithm math is relatively standard, the empirical focus on feature importance and predictive enhancement adds practical utility.

Research Flowchart

  flowchart TD
    A[Research Goal: Evaluate Personalized PageRank PPR<br/>for Fraud Detection] --> B[Data Inputs: Historical Online Transactions<br/>& Account Relationships]
    B --> C[Graph Construction & Calculation<br/>of PPR Network Features]
    C --> D[Model Training & Comparison:<br/>Baseline vs. PPR-Enhanced Models]
    D --> E[Feature Stability & Importance Analysis]
    E --> F[Key Outcome: PPR Features Significantly<br/>Improve Fraud Predictive Power & Model Performance]

Cited References (69)

  1. Beyond Supervised Learning in Credit Card Fraud Detection: A Dive into Semi-supervised and Distributed Learning (2018)
  2. A model based on convolutional neural network for online transaction fraud detection (2018)
  3. Fraud detection and verification system for online transactions: a brief overview (2020)
  4. A survey of feature selection and feature extraction techniques in machine learning (2014)
  5. Improving a Rule-based Fraud Detection System with Classification Based on Association Rule Mining (2021)
  6. Interpretable random forests via rule extraction (2021)
  7. Integrating knowledge-driven and data-driven approaches for the derivation of clinical prediction rules (2005)
  8. Anomaly detection with autoencoder and random forest (2020)
and 61 more
  1. Modular design patterns for hybrid learning and reasoning systems: a taxonomy, patterns and use cases (2021)
  2. Credit card fraud detection using machine learning techniques: A comparative analysis (2017)
  3. Credit card fraud detection with a neural-network (1994)
  4. A new binary support vector system for increasing detection rate of credit card fraud (2006)
  5. Graph-based anomaly detection (2003)
  6. Oddball: Spotting anomalies in weighted graphs (2010)
  7. Credit card fraud detection using Bayesian and neural networks (2002)
  8. Application of classification models on credit card fraud detection (2007)
  9. Association rules applied to credit card fraud detection (2009)
  10. Solving credit card fraud detection problem by the new metaheuristics migrating birds optimization (2013)
  11. Cardwatch: A neural network based database mining system for credit card fraud detection (1997)
  12. Neural data mining for credit card fraud detection (1999)
  13. Parallel granular neural networks for fast credit card fraud detection (2002)
  14. APATE: A novel approach for automated credit card transaction fraud detection using network-based extensions (2015)
  15. Opinion fraud detection in online reviews by network effects (2013)
  16. Neighborhood formation and anomaly detection in bipartite graphs (2005)
  17. Financial Fraud Detection with Graph Data Science (2021) — link
  18. Random forest for credit card fraud detection (2018)
  19. A review of fraud detection techniques: Credit card (2012)
  20. A critical review of credit card fraud detection techniques (2020)
  21. Graph analytics for real-time scoring of cross-channel transactional fraud (2017)
  22. Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation (2020) — link
  23. Accuracy (Error Rate) (2023)
  24. Precision (2023)
  25. Interpretable machine learning with Python: Learn to build interpretable high-performance models with hands-on real-world examples (2021)
  26. Fraudulent transaction detection in credit card by applying ensemble machine learning techniques (2019)
  27. A unified approach to interpreting model predictions (2017)
  28. " Why should i trust you?" Explaining the predictions of any classifier (2016)
  29. Logistic regression (2007)
  30. Node embeddings in social network analysis (2015)
  31. The anatomy of a large-scale hypertextual web search engine (1998)
  32. Topic-sensitive pagerank (2002)
  33. Behavior language processing with graph based feature generation for fraud detection in online lending (2018)
  34. Model Monitoring and Robustness of In-Use Machine Learning Models: Quantifying Data Distribution Shifts Using Population Stability Index (2023) — link
  35. ROC and AUC with a binary predictor: a potentially misleading metric (2020)
  36. Understanding auc-roc curve (2018)
  37. Novel techniques for fraud detection in mobile telecommunication networks (1997)
  38. Discovery of fraud rules for telecommunications—challenges and solutions (1999)
  39. Survey of fraud detection techniques (2004)
  40. The application of data mining techniques in financial fraud detection: A classification framework and an academic review of literature (2011)
  41. Data mining techniques for the detection of fraudulent financial statements (2007)
  42. Auto claim fraud detection using Bayesian learning neural networks (2005)
  43. Assessing the risk of management fraud through neural network technology (1997)
  44. Forecasting fraudulent financial statements using data mining (2006)
  45. A Bayesian dichotomous model with asymmetric link for fraud in insurance (2008)
  46. A case study of applying boosting Naive Bayes to claim fraud diagnosis (2004)
  47. Cost-sensitive payment card fraud detection based on dynamic random forest and k-nearest neighbors (2018)
  48. Credit card fraud detection using anti-k nearest neighbor algorithm (2012)
  49. An intelligent system for user behavior detection in Internet Banking (2015)
  50. A novel idea for credit card fraud detection using decision tree (2017)
  51. Credit card fraud detection through parenclitic network analysis (2018)
  52. A hybrid approach to detect credit card fraud (2015)
  53. Ontology based expert-system for suspicious transactions detection (2014)
  54. Enhancing semantic consistency in anti-fraud rule-based expert systems (2017)
  55. Graph-based fraud detection with the free energy distance (2020)
  56. A graph-based, semi-supervised, credit card fraud detection system (2016)
  57. Random-walk computation of similarities between nodes of a graph with application to collaborative recommendation (2007)
  58. Itemrank: A random-walk based scoring algorithm for recommender engines. (2007)
  59. Evaluating feature importance estimates (2018)
  60. The distribution of PageRank follows a power-law only for particular values of the damping factor (2006)
  61. SHAP and LIME: an evaluation of discriminative power in credit risk (2021)

0 of 69 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.