UPB at SemEval-2020 Task 11: Propaganda Detection with Domain-Specific Trained BERT

TL;DR

This paper presents a domain-specific BERT model trained on propagandistic news to improve detection of propaganda techniques, achieving competitive results in SemEval-2020 tasks.

Contribution

The paper introduces a specialized BERT model trained on propagandistic news for better propaganda detection, outperforming baseline models in key subtasks.

Findings

F1-score of 46.060% in propaganda span identification

F1 score of 54.302% in propaganda technique classification

Ranked 5th and 19th in respective subtasks

Abstract

Manipulative and misleading news have become a commodity for some online news outlets and these news have gained a significant impact on the global mindset of people. Propaganda is a frequently employed manipulation method having as goal to influence readers by spreading ideas meant to distort or manipulate their opinions. This paper describes our participation in the SemEval-2020, Task 11: Detection of Propaganda Techniques in News Articles competition. Our approach considers specializing a pre-trained BERT model on propagandistic and hyperpartisan news articles, enabling it to create more adequate representations for the two subtasks, namely propaganda Span Identification (SI) and propaganda Technique Classification (TC). Our proposed system achieved a F1-score of 46.060% in subtask SI, ranking 5th in the leaderboard from 36 teams and a micro-averaged F1 score of 54.302% for subtask TC, ranking 19th from 32 teams.