Understanding Structured Financial Data with LLMs: A Case Study on Fraud Detection
This addresses the problem of limited interpretability and heavy feature engineering in fraud detection for analysts, though it is incremental as it narrows but does not close the performance gap with specialized classifiers.
The paper tackled fraud detection in financial transactions by introducing FinFRE-RAG, a two-stage LLM-based method that improves F1/MCC scores over direct prompting and is competitive with tabular baselines, while providing interpretable rationales.
Detecting fraud in financial transactions typically relies on tabular models that demand heavy feature engineering to handle high-dimensional data and offer limited interpretability, making it difficult for humans to understand predictions. Large Language Models (LLMs), in contrast, can produce human-readable explanations and facilitate feature analysis, potentially reducing the manual workload of fraud analysts and informing system refinements. However, they perform poorly when applied directly to tabular fraud detection due to the difficulty of reasoning over many features, the extreme class imbalance, and the absence of contextual information. To bridge this gap, we introduce FinFRE-RAG, a two-stage approach that applies importance-guided feature reduction to serialize a compact subset of numeric/categorical attributes into natural language and performs retrieval-augmented in-context learning over label-aware, instance-level exemplars. Across four public fraud datasets and three families of open-weight LLMs, FinFRE-RAG substantially improves F1/MCC over direct prompting and is competitive with strong tabular baselines in several settings. Although these LLMs still lag behind specialized classifiers, they narrow the performance gap and provide interpretable rationales, highlighting their value as assistive tools in fraud analysis.