Events are presented in a carousel, with a main word and other secondary related words describing the event. The full event can be classified as positive or negative by using the red or green buttons:

Below the carousel there are different sections that can be folded, like in the example below. The only unfolded section presents a timeline with the event impact. You can click on some of the words related to the event to hide their impact on the timeline.

For each event, the image clusters are presented below the carousel. You can classify the full cluster by clicking on the positive and negative buttons that appear when you point the image cluster.

You can also explore the n-grams associated to the event. You can filter them according to the tweet label.

If you click on a n-gram bubble, you can access the associated tweets to such n-gram and classify them massively (the buttons at the bottom of the windos) or individually:

The last section presents the individual tweets associated to the event: