Table Caption Generation in Scholarly Documents Leveraging Pre-trained Language Models

TL;DR

This paper presents a method for generating scholarly table captions by retrieving relevant sentences from the paper and using pre-trained language models like T5, demonstrating improved caption quality over GPT-2.

Contribution

It introduces a new dataset, DocBank-TB, and compares different retrieval and generation strategies for scholarly table captioning.

Findings

T5 outperforms GPT-2 in BLEU and METEOR scores

Retrieving relevant sentences improves caption quality

Using sentences matching row headers or the whole table is effective

Abstract

This paper addresses the problem of generating table captions for scholarly documents, which often require additional information outside the table. To this end, we propose a method of retrieving relevant sentences from the paper body, and feeding the table content as well as the retrieved sentences into pre-trained language models (e.g. T5 and GPT-2) for generating table captions. The contributions of this paper are: (1) discussion on the challenges in table captioning for scholarly documents; (2) development of a dataset DocBank-TB, which is publicly available; and (3) comparison of caption generation methods for scholarly documents with different strategies to retrieve relevant sentences from the paper body. Our experimental results showed that T5 is the better generation model for this task, as it outperformed GPT-2 in BLEU and METEOR implying that the generated text are clearer and more precise. Moreover, inputting relevant sentences matching the row header or whole table is effective.