newsSweeper at SemEval-2020 Task 11: Context-Aware Rich Feature Representations For Propaganda Classification

TL;DR

This paper presents a system using enhanced BERT and RoBERTa models with contextual and tagging features for detecting propaganda spans and techniques in news articles, achieving competitive results in SemEval-2020.

Contribution

It introduces a novel combination of pre-trained language models with NER tagging and contextual features for propaganda detection and classification.

Findings

Ranked 5th in propaganda technique classification

Effective use of NER tagging with BERT for span detection

Incorporation of contextual features improves classification accuracy

Abstract

This paper describes our submissions to SemEval 2020 Task 11: Detection of Propaganda Techniques in News Articles for each of the two subtasks of Span Identification and Technique Classification. We make use of pre-trained BERT language model enhanced with tagging techniques developed for the task of Named Entity Recognition (NER), to develop a system for identifying propaganda spans in the text. For the second subtask, we incorporate contextual features in a pre-trained RoBERTa model for the classification of propaganda techniques. We were ranked 5th in the propaganda technique classification subtask.