Does Your Model Classify Entities Reasonably? Diagnosing and Mitigating Spurious Correlations in Entity Typing

TL;DR

This paper investigates biases in entity typing models caused by spurious correlations, systematically defines various bias types, and proposes a counterfactual data augmentation method to improve model robustness and generalization.

Contribution

The paper introduces a comprehensive bias taxonomy in entity typing and presents a counterfactual data augmentation technique to mitigate these biases and enhance model reliability.

Findings

Improved generalization on original and debiased test sets.

Identification of six key types of model biases.

Counterfactual augmentation enhances model understanding.

Abstract

Entity typing aims at predicting one or more words that describe the type(s) of a specific mention in a sentence. Due to shortcuts from surface patterns to annotated entity labels and biased training, existing entity typing models are subject to the problem of spurious correlations. To comprehensively investigate the faithfulness and reliability of entity typing methods, we first systematically define distinct kinds of model biases that are reflected mainly from spurious correlations. Particularly, we identify six types of existing model biases, including mention-context bias, lexical overlapping bias, named entity bias, pronoun bias, dependency bias, and overgeneralization bias. To mitigate model biases, we then introduce a counterfactual data augmentation method. By augmenting the original training set with their debiased counterparts, models are forced to fully comprehend sentences and discover the fundamental cues for entity typing, rather than relying on spurious correlations for shortcuts. Experimental results on the UFET dataset show our counterfactual data augmentation approach helps improve generalization of different entity typing models with consistently better performance on both the original and debiased test sets.