Gender Bias and Universal Substitution Adversarial Attacks on Grammatical Error Correction Systems for Automated Assessment

TL;DR

This paper investigates how gender bias and universal substitution adversarial attacks can deceive grammatical error correction systems, which are used for automated language assessment, by making small input changes that hide errors and falsely improve fluency scores.

Contribution

It introduces a realistic adversarial attack method targeting GEC systems used in automated assessment, highlighting vulnerabilities related to gender bias and attack effectiveness.

Findings

Adversarial attacks can significantly reduce detected errors in GEC outputs.

Gender bias influences the success of adversarial manipulations.

The proposed attack demonstrates potential to deceive GEC-based assessments.

Abstract

Grammatical Error Correction (GEC) systems perform a sequence-to-sequence task, where an input word sequence containing grammatical errors, is corrected for these errors by the GEC system to output a grammatically correct word sequence. With the advent of deep learning methods, automated GEC systems have become increasingly popular. For example, GEC systems are often used on speech transcriptions of English learners as a form of assessment and feedback - these powerful GEC systems can be used to automatically measure an aspect of a candidate's fluency. The count of \textit{edits} from a candidate's input sentence (or essay) to a GEC system's grammatically corrected output sentence is indicative of a candidate's language ability, where fewer edits suggest better fluency. The count of edits can thus be viewed as a \textit{fluency score} with zero implying perfect fluency. However, although deep learning based GEC systems are extremely powerful and accurate, they are susceptible to adversarial attacks: an adversary can introduce a small, specific change at the input of a system that causes a large, undesired change at the output. When considering the application of GEC systems to automated language assessment, the aim of an adversary could be to cheat by making a small change to a grammatically incorrect input sentence that conceals the errors from a GEC system, such that no edits are found and the candidate is unjustly awarded a perfect fluency score. This work examines a simple universal substitution adversarial attack that non-native speakers of English could realistically employ to deceive GEC systems used for assessment.