A Comprehensive Graph Framework for Question Answering with Mode-Seeking Preference Alignment

TL;DR

This paper introduces GraphMPA, a graph-based framework for question answering that improves understanding and aligns responses with human preferences using hierarchical graphs and mode-seeking optimization.

Contribution

It presents a novel hierarchical document graph construction and mode-seeking preference alignment method for enhanced question answering performance.

Findings

Outperforms existing methods on six datasets

Effectively aligns model outputs with human preferences

Improves understanding through hierarchical graph modeling

Abstract

Recent advancements in retrieval-augmented generation (RAG) have enhanced large language models in question answering by integrating external knowledge. However, challenges persist in achieving global understanding and aligning responses with human ethical and quality preferences. To address these issues, we propose GraphMPA, a comprehensive graph-based framework with mode-seeking preference alignment. Our approach constructs a hierarchical document graph using a general similarity measurement, mimicking human cognitive processes for information understanding and synthesis. Additionally, we introduce mode-seeking preference optimization to better align model outputs with human preferences through probability-matching constraints. Extensive experiments on six datasets demonstrate the effectiveness of our \href{https://github.com/tangquanwei/GraphMPA}{GraphMPA}.