Automated Extraction of Records of Processing Activities (RoPA) Using Hybrid RAG and Locally Deployed Large Language Models
Summary
This paper proposes RoPA Manager, an automated system for extracting Records of Processing Activities using hybrid retrieval and locally deployed large language models, evaluated on a Vietnamese benchmark.
View Cached Full Text
Cached at: 09/24/26, 09:19 AM
# Automated Extraction of Records of Processing Activities (RoPA) Using Hybrid RAG and Locally Deployed Large Language Models Source: [https://arxiv.org/abs/2609.27359](https://arxiv.org/abs/2609.27359) [View PDF](https://arxiv.org/pdf/2609.27359) > Abstract:Vietnam's Personal Data Protection Law \(Law No\. 91/2025/QH15\) and Decree No\. 356/2025/ND\-CP, effective January 1, 2026, require organizations to establish and maintain Records of Processing Activities \(RoPA\)\. Manual RoPA preparation is labor\-intensive, while cloud\-hosted large language models \(LLMs\) may conflict with data\-sovereignty requirements\. We propose RoPA Manager, a system for automated RoPA information extraction using hybrid retrieval that combines lexical ranking over tsvector, dense\-vector search, Reciprocal Rank Fusion \(RRF\), and locally deployed LLMs\. We introduce a Vietnamese RoPA benchmark with 32 organizations, 77 processing activities, 12 field groups, and 4,338 reference values\. Evaluation is reported at three distinct levels\. The automated scorer, tested on perturbed data without invoking an LLM, achieved F1 = 0\.9493 \[0\.9436, 0\.9548\]; this measures scorer robustness rather than end\-to\-end extraction accuracy\. End\-to\-end extraction achieved token coverage of 50\.04\-55\.25% against the reference labels\. Two independent experts reviewed 1,558 reference values \(35\.9% of the benchmark\), found no incorrect values, and achieved 99\.68% agreement with PABAK = 0\.9936\. Value\-level precision was not measured\. Across 32 paired scenarios on a 24 GB GPU, locally deployed Qwen3\.5\-27B\-GPTQ\-Int4 showed no statistically significant difference from cloud\-based DeepSeek\-V4\-Flash \(difference 0\.20 percentage points in favor of DeepSeek, 95% CI \[\-0\.93, 1\.32\], p = 0\.72\), while Gemma\-4\-31B performed significantly worse \(p < 0\.01\)\. ## Submission history From: Khuong Nguyen\-An \[[view email](https://arxiv.org/show-email/f5d87041/2609.27359)\] **\[v1\]**Wed, 23 Sep 2026 04:55:53 UTC \(589 KB\)
Similar Articles
HuRo: Robotizing Human Videos for Scalable VLA Pretraining
This paper presents HuRo, a pipeline for robotizing human videos to create scalable VLA pretraining data, showing significant improvements in task completion and robustness on real-world manipulation tasks.
Unleashing the Potential of Large Language Models: A Blueprint for Real-Time, Enterprise-Ready Deployments
This arXiv paper presents a unified LLMOps architecture for real-time, enterprise-ready LLM deployments, integrating data ingestion, continual learning, RAG, and feedback loops. It introduces components like AIPO, STAR+FAR, and SAGE to address knowledge staleness, hallucination, and latency-cost trade-offs in regulated sectors.
Retrieval-Augmented Large Language Models for Schema-Constrained Clinical Information Extraction
This paper presents a modular retrieval-augmented generation (RAG) pipeline for extracting structured clinical observations from conversational nurse-patient transcripts, using schema-constrained prompting and second-pass auditing with Llama and GPT backbones, achieving 80.36% F1 score.
Using Large Language Models to Support High Volume Application Review for an Undergraduate Research Program
This paper describes the development of an LLM-based tool using OpenAI's GPT models to evaluate approximately 1,200 Statements of Purpose for Purdue's SURF program, processing them in 4.6 hours and accelerating the review process compared to traditional human grading.
A Locally Deployed RAG-Based Academic Advising System for Course Selection
This paper proposes a locally deployed RAG-based academic advising system that combines large language models with retrieval from structured syllabus data to support course selection and personalized study planning in a privacy-preserving manner.