Paste a target text (or upload a text file) to analyze the emotion word usage in it. Compare it with a control text of everyday language on social media (or with your own control text). The text is analyzed in your browser and is never uploaded.
A plain text file (one text per line), a CSV or tab-separated file, or a JSON file. A CSV or JSON file needs a text field (also accepted: Tweet) and may also have id (ID, TweetID) and date (createdAt) fields.
The panels, chart and table in this part change with the choice under “Analyze by”. Everything above it is calculated from the whole text.
If you use The NRC Emotion Viz, please cite the paper and the website:
Saif M. Mohammad. TBD (title). TBD (venue).
The NRC Emotion Viz. Website: TBD (URL).
The NRC Emotion Viz draws on the lexicons below. If you report results obtained with a lexicon, cite its paper as well.
More lexicons: saifmohammad.com/WebPages/lexicons.html
When using an off-the-shelf emotion word analysis tool, treat its scores as rough indicators of word usage, not definitive measures of what an author or reader feels. Results depend on the tool’s vocabulary, design, and the choice of target and control texts, so comparisons can mislead if those texts differ in length, genre, context, or language. Avoid using the output to label individuals or make high-stakes decisions, and explain the tool’s limits when sharing findings.
Please see the papers below for ethical considerations involved in automatic emotion detection and the use of emotion lexicons. (These also act as the Ethics and Data Statements for the lexicons.)
Feedback, bug reports and requests for new features, metrics or lexicons are welcome. Email saif.mohammad@nrc-cnrc.gc.ca
Dr. Saif M. Mohammad, National Research Council Canada
saif.mohammad@nrc-cnrc.gc.ca
Beta. The NRC Emotion Viz is still undergoing testing, and its results and features may change. We welcome bug reports and feature requests: see Feedback and Feature Requests above.
The NRC Emotion Viz (Beta) · Page last updated November 2026