Material for the workshop at Appled Machine Learning Days 2019: "Machine Learning for fake news detection: theory and practice." Any content of this repository before the 27th January 2019 is subject to change.
https://github.com/FakeNewsChallenge/fnc-1 Dataset is provided in this repository.
curl https://www.dropbox.com/s/bgfpgko57vicqj2/trainingdatatokens.pkl?dl=0 -L -o trainingdatatokens.pkl
curl https://www.dropbox.com/s/o38caqmlcehgq85/data_tfidf.pkl?dl=0 -L -o data_tfidf.pkl
- Install Anaconda (miniconda would suffice)
- Import the environment from the .yml file:
conda env create -f amld.yml
- Activate the environment
source activate amld
http://www.cits.ucsb.edu/fake-news
Introduction to IR/NLP – https://nlp.stanford.edu/IR-book/information-retrieval-book.html
Introduction to deep learning – https://github.com/janishar/mit-deep-learning-book-pdf
Deep NLP (Stanford) - http://cs224d.stanford.edu/ or http://web.stanford.edu/class/cs224n/
Video Lectures (YouTube) – https://www.youtube.com/playlist?list=PLqdrfNEc5QnuV9RwUAhoJcoQvu4Q46Lja
9:00 - 9:15: Introduction
9:15 - 9:30: Problem statement
9:30 - 9:35: Dataset
9:35 - 10:30: Introduction to Natural Language Processing
10:30 - 10:45: Coffee Break / Technical issues
10:45 - 12:00: Hands on session (introduction to general techniques)
12:00 - 14:00: Lunch break / Technical issues
14:00 - 15:15: Introduction to Neural networks
15:15 - 15:30: Coffee Break / Technical issues
15:30 - end: Discussion / hands on session (modifying/exploring a fake news detection engine)