DiNeR: a Large Realistic Dataset for Evaluating Compositional Generalization

TL;DR

The paper introduces DiNeR, a large realistic Chinese dataset for evaluating compositional generalization through dish name recognition, addressing limitations of synthetic datasets by including diverse linguistic phenomena and real-world complexity.

Contribution

It presents a new large-scale, realistic dataset for dish name recognition that captures diverse linguistic phenomena, along with baseline models and insights into compositional generalization.

Findings

Baseline models show varying performance on the dataset.

The dataset includes complex linguistic phenomena like anaphora and ambiguity.

DiNeR provides a challenging benchmark for compositional generalization.

Abstract

Most of the existing compositional generalization datasets are synthetically-generated, resulting in a lack of natural language variation. While there have been recent attempts to introduce non-synthetic datasets for compositional generalization, they suffer from either limited data scale or a lack of diversity in the forms of combinations. To better investigate compositional generalization with more linguistic phenomena and compositional diversity, we propose the DIsh NamE Recognition (DiNeR) task and create a large realistic Chinese dataset. Given a recipe instruction, models are required to recognize the dish name composed of diverse combinations of food, actions, and flavors. Our dataset consists of 3,811 dishes and 228,114 recipes, and involves plenty of linguistic phenomena such as anaphora, omission and ambiguity. We provide two strong baselines based on T5 and large language models (LLMs). This work contributes a challenging task, baseline methods to tackle the task, and insights into compositional generalization in the context of dish name recognition. Code and data are available at https://github.com/Jumpy-pku/DiNeR.