# Team QCRI-MIT at SemEval-2019 Task 4: Propaganda Analysis Meets Hyperpartisan News Detection

## TL;DR

This paper presents a system for hyperpartisan news detection using propaganda-related features and logistic regression, achieving around 73% accuracy on manually annotated data.

## Contribution

It introduces a feature-based approach leveraging propaganda detection techniques for hyperpartisan news classification.

## Key findings

- Achieved 72.9% accuracy on manually annotated test data.
- Distant supervision test data accuracy was 60.8%.
- Feature pre-processing significantly improves performance.

## Abstract

In this paper, we describe our submission to SemEval-2019 Task 4 on Hyperpartisan News Detection. Our system relies on a variety of engineered features originally used to detect propaganda. This is based on the assumption that biased messages are propagandistic in the sense that they promote a particular political cause or viewpoint. We trained a logistic regression model with features ranging from simple bag-of-words to vocabulary richness and text readability features. Our system achieved 72.9% accuracy on the test data that is annotated manually and 60.8% on the test data that is annotated with distant supervision. Additional experiments showed that significant performance improvements can be achieved with better feature pre-processing.