Narrating Causal Graphs with Large Language Models
This work addresses the challenge of narrating causal graphs for applications like healthcare or marketing, but it is incremental as it builds on existing methods for graph-to-text generation.
The paper tackled the problem of generating text descriptions from causal graphs using large language models, finding that while causal descriptions are harder to generate in zero-shot settings, they improve with training data and can be deployed quickly with few examples.
The use of generative AI to create text descriptions from graphs has mostly focused on knowledge graphs, which connect concepts using facts. In this work we explore the capability of large pretrained language models to generate text from causal graphs, where salient concepts are represented as nodes and causality is represented via directed, typed edges. The causal reasoning encoded in these graphs can support applications as diverse as healthcare or marketing. Using two publicly available causal graph datasets, we empirically investigate the performance of four GPT-3 models under various settings. Our results indicate that while causal text descriptions improve with training data, compared to fact-based graphs, they are harder to generate under zero-shot settings. Results further suggest that users of generative AI can deploy future applications faster since similar performances are obtained when training a model with only a few examples as compared to fine-tuning via a large curated dataset.