GROUNDEDKG-RAG: Grounded Knowledge Graph Index for Long-document Question Answering

2026-04-05

AI Summary

This work addresses the limitations of existing long-document question answering systems based on retrieval-augmented generation (RAG), which often incur high computational costs, generate redundant content, and suffer from hallucinations due to overreliance on large language models for descriptive generation. To mitigate these issues, the authors propose a fully source-anchored, interpretable knowledge graph construction method that represents entities and actions as nodes and temporal or semantic relations as edges. By integrating Semantic Role Labeling (SRL) and Abstract Meaning Representation (AMR) parsing, the approach yields a graph structure that is both human-readable and machine-retrievable. Evaluated on the NarrativeQA dataset, the method achieves factual accuracy and efficiency comparable to state-of-the-art proprietary long-context models at lower cost, outperforms existing baselines, and enables auditability of generated answers.

Technology Category

• Natural Language Processing: Question Answering
• Knowledge Representation and Reasoning: Knowledge Representation Languages
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Retrieval-augmented generation (RAG) systems have been widely adopted in contemporary large language models (LLMs) due to their ability to improve generation quality while reducing the required input context length. In this work, we focus on RAG systems for long-document question answering. Current approaches suffer from a heavy reliance on LLM descriptions resulting in high resource consumption and latency, repetitive content across hierarchical levels, and hallucinations due to no or limited grounding in the source text. To improve both efficiency and factual accuracy through grounding, we propose GroundedKG-RAG, a RAG system in which the knowledge graph is explicitly extracted from and grounded in the source document. Specifically, we define nodes in GroundedKG as entities and actions, and edges as temporal or semantic relations, with each node and edge grounded in the original sentences. We construct GroundedKG from semantic role labeling (SRL) and abstract meaning representation (AMR) parses and then embed it for retrieval. During querying, we apply the same transformation to the query and retrieve the most relevant sentences from the grounded source text for question answering. We evaluate GroundedKG-RAG on examples from the NarrativeQA dataset and find that it performs on par with a state-of-the art proprietary long-context model at smaller cost and outperforms a competitive baseline. Additionally, our GroundedKG is interpretable and readable by humans, facilitating auditing of results and error analysis.

Problem

• retrieval-augmented generation
• long-document question answering
• hallucination
• grounding
• knowledge graph

Innovation

• Grounded Knowledge Graph
• Retrieval-Augmented Generation
• Semantic Role Labeling
• Abstract Meaning Representation
• Long-document Question Answering