Evaluating Generative Models for Graph-to-Text Generation
This work addresses the resource-intensive finetuning problem for graph-to-text generation, though it is incremental as it primarily benchmarks existing models.
The paper evaluated GPT-3 and ChatGPT for graph-to-text generation in a zero-shot setting, finding they achieved BLEU scores of 10.57 and 11.08 on AGENDA and WebNLG datasets but struggled with semantic relations and hallucinations.
Large language models (LLMs) have been widely employed for graph-to-text generation tasks. However, the process of finetuning LLMs requires significant training resources and annotation work. In this paper, we explore the capability of generative models to generate descriptive text from graph data in a zero-shot setting. Specifically, we evaluate GPT-3 and ChatGPT on two graph-to-text datasets and compare their performance with that of finetuned LLM models such as T5 and BART. Our results demonstrate that generative models are capable of generating fluent and coherent text, achieving BLEU scores of 10.57 and 11.08 for the AGENDA and WebNLG datasets, respectively. However, our error analysis reveals that generative models still struggle with understanding the semantic relations between entities, and they also tend to generate text with hallucinations or irrelevant information. As a part of error analysis, we utilize BERT to detect machine-generated text and achieve high macro-F1 scores. We have made the text generated by generative models publicly available.