News Audit

Abstract

1. Introduction

This project is archived and will not receive further updates, bug fixes, or security patches. Issues and pull requests may not be reviewed.

The first step of this project was creating a database of news domains. I knew of several resources that publish information about news domains - for example, OpenSources (http://www.opensources.co/), a curated resource for assessing online information sources, uses a combination of a dozen tags (such as ‘fake news’, ‘satire’, ‘extreme bias’ etc.) to assess domains, and this information is available for public use. In addition, Guy DePauw sent me a similar list of sources and tags from PolitiFact, though the tags were slightly different. Finally, Tom De Smedt pointed me to MediaBiasFactCheck.com, another resource that contains 1600+ media sources and categorizes them into classes such as ‘questionable’, ‘right’, right-center’, ‘least-biased’, ‘left, ‘left-center’, ‘conspiracy’, ‘pseudoscience’, etc.

Naturally, there was a bit of overlap between these three resources, and while having information from more than one of them would be helpful, I still wanted a centralized database for all of the domains. The challenge was that the three resources used different categorization schemas and sometimes referred to the same domains differently (for example, some only had the URL and not the name, some only the name and not the URL, and even then it wasn’t always an exact match (ex. www.cnn.com vs cnn.com, The Daily Buzz vs. daily buzz). I normalized the URLs and used various string similarity/distance measures (such as Levenshtein distance) to group together duplicates. In the end, I aggregated all of this categorization in one place and put together a CSV of all domains from the three sources above (~2k domains), along with the categories assigned by each, any additional comments, etc. For those sources that lacked URL information, I made sure to get that information by running the name of the source through the Google API - typically the first returned result was the correct URL, but I did check double-check this manually.

2. Objective

Fake news detection, Google Summer of Code 2017

This project demonstrates how Python can be applied to a real-world AI & Machine Learning problem.

4. Technology Stack

5. System Requirements

General requirements for this technology stack — check the README for exact versions.

• Python 3.8 or later
• pip / virtualenv for dependencies
• VS Code, PyCharm or Jupyter Notebook
• Git (to clone the repository)

6. Installation & Setup

git clone https://github.com/clips/news-audit.git
cd news-audit

Full setup instructions are in the project README.

7. Future Enhancements

Suggested extensions you can add to make this your own project.

• Deploy the model as a web app with Streamlit, Flask or FastAPI
• Compare against an additional model and report the metric difference
• Add explainability (SHAP / Grad-CAM)

8. Viva / Review Questions

Common questions examiners ask for projects in this domain.

1. What dataset does the project use and how was it pre-processed?
2. Which algorithm / model architecture is used and why was it chosen over alternatives?
3. How are training and testing data split, and how is overfitting avoided?
4. Which evaluation metrics (accuracy, precision, recall, F1) are reported and what do they mean here?
5. How would you deploy this model for real users?

9. Source Code & License

This project is developed by clips and published on GitHub under the GNU General Public License v3.0. Please follow the license terms and credit the original author when you use or modify this code.