Context Based on the high rate of cyber-bulling and other critical bad behaviors on internet is the need to anticipate and predict if any user need help. Content Here we collect diferents history of tweets of users (normal and kind of sad) so we can investigate if there is any pattern and predict the user emotion of that day. We create the score of each emotion based on the Emolex (https://saifmohammad.com/WebPages/NRC-Emotion-Lexicon.htm) which is a lexic tag based on the Plutchick wheel of emotions. ![enter image description here][2] Acknowledgements We manually collect this dataset thanks to the API of twitter for a course project. We implement a predictive model and we are happy a bout it. * Margarita Bugue?o * Fabián Fernandez * Francisco Mena We also want to thanks to Saif Mohammad who provided the emolex tags, and thanks to the National Research Council Canada. Inspiration The data here can be used to understand and identify emotions reflected in the tweets and also Answer calls that present levels critics of anguish or annoyance. [2]: https://upload.wikimedia.org/wikipedia/commons/thumb/c/ce/Plutchik-wheel.svg/2000px-Plutchik-wheel.svg.png "emotion"