According to a paper published on arxiv.org, researchers have introduced MATRAG (Multi-Agent Transparent Retrieval-Augmented Generation), a framework designed to address transparency challenges in Large Language Model (LLM)-based recommendation systems.
MATRAG employs four specialized agents: a User Modeling Agent that constructs dynamic preference profiles, an Item Analysis Agent that extracts semantic features from knowledge graphs, a Reasoning Agent that synthesizes collaborative and content-based signals, and an Explanation Agent that generates natural language justifications grounded in retrieved knowledge, according to the paper.
The framework incorporates “a transparency scoring mechanism that quantifies explanation faithfulness and relevance,” the paper states. In experiments on three benchmark datasets (Amazon Reviews, MovieLens-1M, and Yelp), MATRAG improved recommendation accuracy by 12.7% in Hit Rate and 15.3% in NDCG compared to leading baselines, according to the research.
Human evaluation confirmed that “87.4% of generated explanations are rated as helpful and trustworthy by domain experts,” the paper reports.
The work appears alongside other recent multi-agent research published on arxiv.org the same day, including KompeteAI for AutoML systems, an empirical comparison of agent communication protocols, and FairQE, a framework for mitigating gender bias in translation quality estimation that was accepted to ACL 2026.