Asosiy kontentga oʻtish
Boshqa

GreenPhish-XAI: Reproducibility Materials for Sustainable and Explainable Human- and LLM-Generated Phishing Detection

Kirti SethInha University in Tashkentseth ashishInha University in Tashkent
2026en
Ochiq kirishYashil

Litsenziya: cc-by

ABI

Annotatsiya

This repository provides reproducibility materials associated with the study “GreenPhish-XAI: A Pareto-Optimized Sustainable and Explainable Framework for Detecting Human- and LLM-Generated Phishing Emails.” The repository contains the final 45-feature derived dataset, feature-selection results, MOPSO operating-point feature subsets, strict leave-one-LLM-out evaluation results, sustainability profiling outputs, explainability analyses, Pareto operating-point results, selected figures, and supporting code where available. The study evaluates phishing detection across benign, human-generated phishing, and phishing generated by GPT-4.1, DeepSeek 3.2, and Llama 3.3 70B. Original third-party email corpora are not redistributed; users should obtain those datasets from their official repositories and comply with the applicable licensing and usage conditions.

Hali tarjima qilinmagan

Identifikatorlar

Iqtiboslar va manbalar

0 ta iqtibos0 ta foydalanilgan manba