Propaganda Detection — Propaganda and Manipulation Detection
Also known as: propaganda and manipulation detection, propaganda technique detection, Propaganda ve Manipülasyon Tespiti
Propaganda detection is a natural-language-processing task that automatically identifies and labels persuasion and manipulation techniques in text — such as loaded language, oversimplified solutions, bandwagon appeals, and glittering generalities. It builds on the fine-grained propaganda analysis introduced by Da San Martino et al. (2019), turning rhetorical manipulation into structured, technique-level labels.
Key highlights
- Produces fine-grained, technique-level labels rather than a single overall polarity score.
- Turns rhetorical manipulation into structured, auditable signals tied to specific spans or documents.
- Builds on an established propaganda-technique taxonomy from the fine-grained analysis literature.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Propaganda detection fits when you have text data and want to flag specific persuasion or manipulation techniques rather than only overall tone. It is a supervised task, so it assumes a labelled training set exists and that a propaganda-technique taxonomy has been defined in advance; aim for at least roughly fifty annotated documents. Without a fixed taxonomy and labelled examples, the model has nothing to learn from.
Strengths & limitations
- Produces fine-grained, technique-level labels rather than a single overall polarity score.
- Turns rhetorical manipulation into structured, auditable signals tied to specific spans or documents.
- Builds on an established propaganda-technique taxonomy from the fine-grained analysis literature.
- Supervised: requires a labelled data set annotated against the chosen taxonomy.
- Results depend on first defining a clear and complete propaganda-technique taxonomy.
- Technique labels are often imbalanced and context-dependent, making some categories hard to detect reliably.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How is this different from sentiment analysis?
Sentiment analysis assigns an overall polarity — positive, negative, or neutral — to text. Propaganda detection instead identifies specific persuasion techniques such as loaded language, bandwagon, or glittering generalities, often at the level of individual spans, so the output is a set of technique labels rather than a single tone.
Do I need labelled data?
Yes. Propaganda detection is a supervised task, so it assumes a labelled training set annotated against your chosen technique taxonomy. The source recommends at least around fifty annotated documents, and without labelled examples the model has nothing to learn from.
What is the propaganda-technique taxonomy and why must it come first?
The taxonomy is the catalogue of techniques to detect — for example loaded language, oversimplification, bandwagon, and glittering generalities. It defines the label set the model learns and predicts, so it must be fixed before data is annotated; otherwise labels are inconsistent and evaluation is unreliable.
How should I evaluate the model?
Score predictions against labelled data and inspect misclassifications. Because technique labels are usually imbalanced, rely on per-class precision, recall, and F1 rather than overall accuracy alone.
Sources
- 1.Da San Martino, G. et al. (2019). Fine-Grained Analysis of Propaganda in News Articles. EMNLP.
- 2.Rashkin, H. et al. (2017). Truth of Varying Shades: Analyzing Language in Fake News and Political Fact-Checking. EMNLP.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 1). Propaganda Detection. ScholarGate. https://scholargate.app/text-mining/propaganda-detection