Who Speaks Matters: Authority-Aware Multi-View RAG over Italian Parliamentary Proceedings

2026-08-13

AI Summary

This work addresses the challenges posed by the vast scale and fragmented nature of parliamentary records, which hinder multi-perspective access, as well as limitations of conventional retrieval-augmented generation (RAG) systems in political contexts—such as dominance by high-frequency speakers, lack of expert weighting, and citation inaccuracies. To overcome these issues, the authors propose ParliamentRAG, a retrieval-augmented generation system tailored to the Italian Chamber of Deputies. It introduces a query-driven topical authority model that dynamically evaluates speaker credibility by integrating interpretable features including professional background, education, and historical contributions. This enables the retrieval and synthesis of cross-party expert viewpoints to generate citation-faithful summaries. Experiments across 15 policy topics demonstrate that ParliamentRAG achieves a political group coverage of 0.97 (versus 0.95 for NotebookLM), perfect citation faithfulness (1.00 vs. 0.95), and significantly higher source relevance as judged by domain experts.

Abstract

Parliamentary proceedings are a primary record of democratic deliberation, yet their volume and fragmentation make multi-perspective access difficult for citizens, journalists, and researchers. Applying Retrieval-Augmented Generation (RAG) to parliamentary transcripts introduces three specific risks: dominance of the most frequent speakers, inability to weight speakers according to topical expertise, and citation misattribution in politically sensitive text. We present ParliamentRAG, a RAG system for the Italian Chamber of Deputies that addresses these risks jointly. Its core contribution is a topic-dependent authority model that estimates each speaker's authority as a function of the current query, combining interpretable components such as profession, education, and previous interventions. Given a user query, the system retrieves relevant speech chunks, identifies topic-relevant experts across parliamentary groups, and generates a summary synthesizing their perspectives, accompanied by supporting quotations. ParliamentRAG is evaluated against Google NotebookLM on 15 policy topics via a two-level protocol combining automated metrics and blind A/B human evaluation by six domain experts. The system achieves higher coverage across political groups (0.97 vs. 0.95), perfect quotation faithfulness (1.00 vs. 0.95), and stronger expert preferences on source-related dimensions, while NotebookLM remains stronger on prose-oriented dimensions.

Problem

• parliamentary proceedings
• Retrieval-Augmented Generation
• authority-aware
• multi-view
• citation misattribution

Innovation

• authority-aware RAG
• topic-dependent authority
• parliamentary proceedings
• retrieval-augmented generation
• expert attribution