ebtardb1's Introduction

Event-Based-Time-Aware-Rumor-Detection-Benchmark

This repository is the official implementation of [Event-Based Time-Aware Rumor Detection Benchmark].

Requirements

To install requirements:

conda env create -f environment.yaml

Training

To train the model(s) in the paper, run this command:


BERT
cd Improving-Rumor-Detection-with-User-Comments-main
cd PHEME-RNR
python Rumor_BERT_domain_discriminator_data.py train event
cd TWITTER-RNR
python Rumor_BERT_domain_discriminator_data.py train event Twitter15

EANN
cd EANN-KDD18-master
cd src
python EANN_text_data.py train event
python EADD_text_twitter_data.py train event Twitter15

GACL
cd GACL-CADA
cd PHEME_two_classification
python main_pheme_domain_discriminator_data.py train b
cd Twitter_four_classification
python main_twitter_domain_discriminator_data.py train b Twitter15

MetaDetector
cd metadetector-master
cd src
python MetaDetector_data.py train event
python MetaDetector_Twitter_data.py train event Twitter15

Pre-trained Models

Our pretrained models here: BERT Download from 'https://www.aliyundrive.com/s/S9RLpi5fERM' and place 'pheme' in BERT into Improving-Rumor-Detection-with-User-Comments-main/PHEME-RNR/model_all_domain Download from 'https://www.aliyundrive.com/s/S9RLpi5fERM' and place 'twitter' in BERT into Improving-Rumor-Detection-with-User-Comments-main/Twitter-RNR/model_all_domain

EANN Download from 'https://www.aliyundrive.com/s/S9RLpi5fERM' and place the files in EANN into EANN-KDD18-master/data/output/ EANN-KDD18-master/data/output/

GACL GACL-CADA/PHEME_two_classification/model_all_domain GACL-CADA/Twitter_four_classification/model_all_domain

MetaDetector metadetector-master/data/output/