An online portal that conducts sentiment analyses of tweets.
We Feel is a tool developed by Australia's national science agency CSIRO that mines Twitter data and maps tweets according to a crowdsourced vocabulary of emotions. Once an emotion is detected, We Feel applies another filter to determine its intensity. This filter is derived from a dataset of 14,000 affective norms provided by the Center for Reading Research at Ghent University, and maps each tweet "against three dimensions: valence (the pleasantness of the stimulus), arousal (the intensity of emotion provoked by the stimulus), and dominance (the degree of control exerted by the stimulus)".
We Feel offers two interfaces from which users can conduct a sentiment analysis - that is, an interactive graphic visualization, or a table view. In both of these views, users can apply a number of filters, such as specific emotions, time ranges, locations and so on. The platform also has an API that lets you draw out and mash-up We Feel's data by emotion, the gender of the tweeter, and originating time zone.
Image: A screenshot of aggregate love tweets by males, here showing that love was the dominant emotion of 1.73% of tweets by males.
Case study: Sydney Siege
Their sentiment analysis on We Feel found that "while there was a surge in emotions such as sadness as expected (see Fig 2), there was also a surge in joy (shown in yellow in Fig 1)."
Image: Wan & Paris, CSIRO
The researchers then coupled We Feel's emotional analysis with a content analysis to find out why these two conflicting emotions concurrently trended. In doing so, they revealed that, whilst the event itself was sad, it also triggered "messages of hope, praises for the police, prayers and unity from a wide spectrum of the community".
Read the full CSIRO study on the Sydney Siege here.
Visit the We Feel platform here.