/NLP-Attack-LocalSearch

Efficient Combinatorial Optimization for Adversarial Textual Attack

Primary LanguagePython

Efficient Combinatorial Optimization for Adversarial Textual Attack

This repository contains source code for the research work described in our TASL paper: Efficient Combinatorial Optimization for Adversarial Textual Attack

Attacking

The folders GA_related, PSO_related, SbGS_related and SGS_related all contain a readme file, which is a step-by-step instruction to obtain the dataset and victim model, run LS and the baseline algorithm, and gather results under the corresponding attack scenarios.