Inno at SemEval-2020 Task 11: Leveraging Pure Transformer for Multi-Class Propaganda Detection

TL;DR

This paper explores using a pure Transformer model with optimized training to classify propaganda techniques in news articles, achieving promising F1 scores across multiple classes.

Contribution

It introduces a Transformer-based approach specifically optimized for multi-class propaganda detection in news articles.

Findings

Achieved 0.6 F1 on validation set

Achieved 0.58 F1 on test set

Non-zero F1 scores for all classes

Abstract

The paper presents the solution of team "Inno" to a SEMEVAL 2020 task 11 "Detection of propaganda techniques in news articles". The goal of the second subtask is to classify textual segments that correspond to one of the 18 given propaganda techniques in news articles dataset. We tested a pure Transformer-based model with an optimized learning scheme on the ability to distinguish propaganda techniques between each other. Our model showed 0.6 and 0.58 overall F1 score on validation set and test set accordingly and non-zero F1 score on each class on both sets.