D-NEXUS: Defending Text Networks using Summarization

Codes and implementation files to be used as a reference to the paper: D-NEXUS: Defending Text Networks using Summarization

Status - Published in Elsevier's Electronic Commerce Research and Applications

Overview

This repository contains the following files:

• Attack_Data - Contains data obtained by running run_attack.sh over the datasets for different models.
• run_attack.sh - For generating attack files using textattack library.
• Summarizer.py - For running proposed defense on attacked data.
• Timer_Summ.py - For inference time comparision with and without the application of the proposed defense.

Further details about the methodology may be directly referred to from the published study.

Citation

If you intend to use this work, kindly cite us as follows:

@article{GUPTA2022DEFENDING,
title = {D-NEXUS: Defending Text Networks Using Summarization},
journal = {Electronic Commerce Research and Applications},
pages = {101171},
year = {2022},
issn = {1567-4223},
doi = {https://doi.org/10.1016/j.elerap.2022.101171},
url = {https://www.sciencedirect.com/science/article/pii/S1567422322000552},
author = {Anup Kumar Gupta and Aryan Rastogi and Vardhan Paliwal and Fyse Nassar and Puneet Gupta},
keywords = {Sentiment Analysis, Natural Language Processing, Adversarial Defenses, Transformers, Adversarial Attack, Language Summarization},
}