PRTA

A Tool For the Analysis of Propaganda Techniques in Texts

This page shows our tool for detecting persuasion techniques in texts.

This demo has two main functionalities:

1. Propaganda technique analysis on a topic

Collecting articles about a topic and showing aggregated statistics on the propaganda techniques our learning algorithm detects in them.
It further allows the user to customise the plots by filtering by source, date, keywords, political bias.
We apply our system to news articles, continuously collected from more than 2K sources.

Go to DEMO

Examples of topics include:

2. Highlighting of the propaganda techniques in a text

The demo shows the list of articles related to a topic per news outlets. When clicking on an article, it shows the spans in the articles in which each technique occurs. Furthermore, we allow users to submit any text and have it analysed by our system

Submit Text for Analysis

References

The model behind the demo is also available as an API.

The demo has been published at ACL 2020:

G. Da San Martino, S. Shaar, Y. Zhang, S. Yu, A. Barrón-Cedeño, P. Nakov,  Prta: A System to Support the Analysis of Propaganda Techniques in the News.   In Proceedings of the 2020 Annual Conference of the Association for Computational Linguistics (ACL 2020), Seattle, USA, July 5-10, 2020.

The learning algorithm that is used to make predictions for this demo is described in the following paper:

G. Da San Martino, S. Yu, A. Barrón-Cedeño, R. Petrov, P. Nakov,  Fine-Grained Analysis of Propaganda in News Articles.   In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP 2019), Hong Kong, China, November 3-7, 2019.

Copyright © QCRI