Generative Entity Typing with Curriculum Learning

TL;DR

This paper introduces a generative entity typing approach using curriculum learning to improve type assignment, especially for long-tail and zero-shot types, outperforming existing models across multiple languages and tasks.

Contribution

It proposes a novel generative paradigm for entity typing combined with curriculum learning to handle data heterogeneity and improve type granularity.

Findings

Outperforms state-of-the-art entity typing models.

Effective in few-shot and zero-shot scenarios.

Applicable across multiple languages and downstream tasks.

Abstract

Entity typing aims to assign types to the entity mentions in given texts. The traditional classification-based entity typing paradigm has two unignorable drawbacks: 1) it fails to assign an entity to the types beyond the predefined type set, and 2) it can hardly handle few-shot and zero-shot situations where many long-tail types only have few or even no training instances. To overcome these drawbacks, we propose a novel generative entity typing (GET) paradigm: given a text with an entity mention, the multiple types for the role that the entity plays in the text are generated with a pre-trained language model (PLM). However, PLMs tend to generate coarse-grained types after fine-tuning upon the entity typing dataset. Besides, we only have heterogeneous training data consisting of a small portion of human-annotated data and a large portion of auto-generated but low-quality data. To tackle these problems, we employ curriculum learning (CL) to train our GET model upon the heterogeneous data, where the curriculum could be self-adjusted with the self-paced learning according to its comprehension of the type granularity and data heterogeneity. Our extensive experiments upon the datasets of different languages and downstream tasks justify the superiority of our GET model over the state-of-the-art entity typing models. The code has been released on https://github.com/siyuyuan/GET.