To Drop or Not to Drop? Predicting Argument Ellipsis Judgments: A Case Study in Japanese

TL;DR

This study investigates how native Japanese speakers decide on argument omission in sentences, providing a large annotated dataset and analyzing the performance of language models in predicting these judgments.

Contribution

It introduces a large-scale annotated dataset of Japanese argument ellipsis judgments and evaluates language models' ability to predict human decisions, revealing linguistic factors involved.

Findings

Native speakers share common criteria for ellipsis decisions.

Language models show gaps compared to human judgments in specific linguistic aspects.

Quantitative analysis of linguistic factors influencing ellipsis judgments.

Abstract

Speakers sometimes omit certain arguments of a predicate in a sentence; such omission is especially frequent in pro-drop languages. This study addresses a question about ellipsis -- what can explain the native speakers' ellipsis decisions? -- motivated by the interest in human discourse processing and writing assistance for this choice. To this end, we first collect large-scale human annotations of whether and why a particular argument should be omitted across over 2,000 data points in the balanced corpus of Japanese, a prototypical pro-drop language. The data indicate that native speakers overall share common criteria for such judgments and further clarify their quantitative characteristics, e.g., the distribution of related linguistic factors in the balanced corpus. Furthermore, the performance of the language model-based argument ellipsis judgment model is examined, and the gap between the systems' prediction and human judgments in specific linguistic aspects is revealed. We hope our fundamental resource encourages further studies on natural human ellipsis judgment.