Cross-Domain Data Selection and Augmentation for Automatic Compliance Detection

TL;DR

This paper investigates data selection methods to improve cross-domain transfer in automated compliance detection, demonstrating that targeted data selection reduces negative transfer and enhances scalability.

Contribution

It systematically evaluates four data selection approaches to mitigate negative transfer in compliance detection framed as an NLI task, advancing cross-domain adaptation techniques.

Findings

Targeted data selection significantly reduces negative transfer.

Embedding-based retrieval improves cross-domain adaptation.

Varying data proportions affects transfer effectiveness.

Abstract

Automating the detection of regulatory compliance remains a challenging task due to the complexity and variability of legal texts. Models trained on one regulation often fail to generalise to others. This limitation underscores the need for principled methods to improve cross-domain transfer. We study data selection as a strategy to mitigate negative transfer in compliance detection framed as a natural language inference (NLI) task. Specifically, we evaluate four approaches for selecting augmentation data from a larger source domain: random sampling, Moore-Lewis's cross-entropy difference, importance weighting, and embedding-based retrieval. We systematically vary the proportion of selected data to analyse its effect on cross-domain adaptation. Our findings demonstrate that targeted data selection substantially reduces negative transfer, offering a practical path toward scalable and reliable compliance automation across heterogeneous regulations.