This repo is implementation of "Fake News Detection Framework Using Word Networks and Propagation Patterns of News in Social Media. Jongmo Kim, Hanmin Kim, Jinuk Cho, Mye Sohn. PROCEEDING OF BIG DATA APPLICATIONS AND SERVICES, Vol.7, No.1, 2019".
We used a dataset based on the repository BuzzFace link. Most of the news stories were no longer found, so we used our own crawler to get the body of the news in the link that still existed.
-
Python 3.6.7
-
nltk 3.2.5
-
sklearn 1.2.0
-
pandas 0.24.2
-
numpy 1.16.4
-
networkx 2.3
in python code:
import nltk
# nltk.download('wordnet')
# nltk.download('punkt')
# nltk.download('averaged_perceptron_tagger')
# nltk.download('stopwords')
If you want to test the code with small dataset (size 4), you need do minor change
python run.py