arXiv · · Significant research
Researchers developed a domain-adapted Retrieval-Augmented Generation (RAG) pipeline to automate regulatory compliance question answering in financial services. The pipeline features a three-stage retriever built on LegalBERT and a compact generator, achieving improved Recall@10 and answer quality scores on ObliQA, a benchmark derived from Abu Dhabi Global Market (ADGM) rulebooks. While showing strong performance for ADGM-specific questions, the adapted models did not transfer to Australian case-law questions, and the study highlighted limitations in evaluating answer grounding. Why it matters: This research offers a practical AI solution for a critical challenge faced by financial institutions in the GCC, demonstrating how AI can enhance efficiency and accuracy in navigating complex regional regulatory frameworks.
Retrieval-Augmented Generation · Financial Services · Regulatory Compliance · Abu Dhabi Global Market · Question Answering
Get the weekly digest
Top AI stories from the GCC region, every week.
arXiv ·
Researchers introduce a new task for generating question-passage pairs to aid in developing regulatory question-answering (QA) systems. The ObliQA dataset, comprising 27,869 questions from Abu Dhabi Global Markets (ADGM) financial regulations, is presented. A baseline Regulatory Information Retrieval and Answer Generation (RIRAG) system is designed and evaluated using the RePASs metric.
arXiv ·
This paper introduces a Regulatory Knowledge Graph (RKG) for the Abu Dhabi Global Market (ADGM) regulations, constructed using language models and graph technologies. A portion of the regulations was manually tagged to train BERT-based models, which were then applied to the rest of the corpus. The resulting knowledge graph, stored in Neo4j, and code are open-sourced on GitHub to promote advancements in compliance automation.
arXiv ·
The QU-NLP team presented their approach to the QIAS 2025 shared task on Islamic Inheritance Reasoning, fine-tuning the Fanar-1-9B model using LoRA and integrating it into a RAG pipeline. Their system achieved an accuracy of 0.858 on the final test, outperforming models like GPT 4.5, LLaMA, and Mistral in zero-shot settings. The system particularly excelled in advanced reasoning, achieving 97.6% accuracy. Why it matters: This demonstrates the effectiveness of domain-specific fine-tuning and retrieval augmentation for Arabic LLMs in complex reasoning tasks, even surpassing frontier models.
arXiv ·
This paper introduces Cross-Document Topic-Aligned (CDTA) chunking to address knowledge fragmentation in Retrieval-Augmented Generation (RAG) systems. CDTA identifies topics across documents, maps segments to topics, and synthesizes them into unified chunks. Experiments on HotpotQA and UAE legal texts show that CDTA improves faithfulness and citation accuracy compared to existing chunking methods, especially for complex queries requiring multi-hop reasoning.