CREFT: Sequential Multi-Agent LLM for Character Relation Extraction

TL;DR

CREFT is a novel multi-agent LLM framework that effectively extracts and refines complex character relations in narratives, significantly improving accuracy and completeness over traditional single-agent methods.

Contribution

It introduces a sequential multi-agent LLM approach for character relation extraction, with a novel knowledge distillation process and iterative refinement, advancing narrative analysis capabilities.

Findings

Outperforms single-agent LLM baselines in accuracy and completeness

Effectively constructs and visualizes character networks

Accelerates script review and narrative comprehension

Abstract

Understanding complex character relations is crucial for narrative analysis and efficient script evaluation, yet existing extraction methods often fail to handle long-form narratives with nuanced interactions. To address this challenge, we present CREFT, a novel sequential framework leveraging specialized Large Language Model (LLM) agents. First, CREFT builds a base character graph through knowledge distillation, then iteratively refines character composition, relation extraction, role identification, and group assignments. Experiments on a curated Korean drama dataset demonstrate that CREFT significantly outperforms single-agent LLM baselines in both accuracy and completeness. By systematically visualizing character networks, CREFT streamlines narrative comprehension and accelerates script review -- offering substantial benefits to the entertainment, publishing, and educational sectors.