Beyond LLMs: A Linguistic Approach to Causal Graph Generation from Narrative Texts
This work addresses the challenge of capturing nuanced causal chains in narratives for researchers and practitioners in natural language processing, offering an interpretable and efficient tool, though it is incremental as it builds on existing LLM and classification methods.
The paper tackles the problem of generating causal graphs from narrative texts by proposing a hybrid framework that combines LLM-based summarization with a linguistically informed Expert Index and STAC classification, achieving superior precision compared to pure LLM approaches like GPT-4o and Claude 3.5 in experiments on 100 narrative chapters and short stories.
We propose a novel framework for generating causal graphs from narrative texts, bridging high-level causality and detailed event-specific relationships. Our method first extracts concise, agent-centered vertices using large language model (LLM)-based summarization. We introduce an "Expert Index," comprising seven linguistically informed features, integrated into a Situation-Task-Action-Consequence (STAC) classification model. This hybrid system, combining RoBERTa embeddings with the Expert Index, achieves superior precision in causal link identification compared to pure LLM-based approaches. Finally, a structured five-iteration prompting process refines and constructs connected causal graphs. Experiments on 100 narrative chapters and short stories demonstrate that our approach consistently outperforms GPT-4o and Claude 3.5 in causal graph quality, while maintaining readability. The open-source tool provides an interpretable, efficient solution for capturing nuanced causal chains in narratives.