Narrative Knowledge Weaver: Narrative-Centric Retrieval-Augmented Reasoning for Long-Form Text Understanding

2026-06-04

AI Summary

This work addresses the challenge of long-form narrative question answering, which requires modeling dynamically evolving storylines and the functional roles of supporting evidence—capabilities inadequately supported by existing methods. The authors propose the Narrative Knowledge Weaver framework, which uniquely aligns multidimensional narrative elements, including textual evidence, atomic facts, and plotlines. By integrating retrieval-augmented generation, knowledge graphs, entity profiling, plot segmentation, and a narrative-specific reader, the framework introduces fine-grained reasoning constraints—such as character roles, temporal order, and state transitions—after retrieval. Evaluated on the STAGE, FairytaleQA, and QuALITY benchmarks, the approach significantly advances script-level question answering performance, demonstrating particularly strong gains in tasks involving character, causal, and temporal reasoning.

Abstract

Long-form narrative QA requires reasoning over evolving story worlds rather than isolated passages: answers may depend on earlier goals, changing character states, social relations, causal triggers, temporal position, and later consequences. Existing retrieval and graph-augmented generation methods improve evidence access, but their units--chunks, entities, relations, summaries, or tool actions--do not directly encode how evidence functions in a story. We introduce Narrative Knowledge Weaver(NKW), a source-grounded framework that aligns textual evidence, atomic facts, canonical graph structure, entity profiles, interactions, episodes, and storylines. At query time, NKW uses text, graph, and narrative tools with post-retrieval reading skills to assemble evidence and audit actor, scope, polarity, state, and temporal constraints. Across STAGE, FairytaleQA, and QuALITY, NKW is strongest on screenplay-level story-world QA while remaining competitive on more passage-centered benchmarks. Ablations, question-type analyses, graph-asset statistics, and case studies show complementary benefits for character, scene, temporal, causal, and narrative-progression reasoning.

Problem

• narrative reasoning
• long-form QA
• story world understanding
• temporal reasoning
• causal reasoning

Innovation

• narrative-centric reasoning
• retrieval-augmented generation
• story-world understanding
• graph-augmented QA
• temporal and causal reasoning