The mood on Dutch Twitter

We sample and analyse a portion of Dutch Twitter traffic continuously. We register top trends (the labels displayed under the horizontal axis) and dominant emotions. We only analyse the text in relation to hashtags and do not process any information pertaining to individual Twitter users.

Our emotion detection service can be used with arbitrary sources of data to plot these and other such charts. As we found out, expressions of happiness seem to signify the prevailing mood on Dutch Twitter. It starts to get more interesting when we zoom in on the trend which generated the most excitement for the past 24 hours. The pie chart below shows the relative distribution of emotions for this trend. The chart is redrawn every whole hour.

To further explore our data, we can lookup the trend which generated the most negative responses with a simple sort operation. The chart below is the result. The emotion ‘surprise’ is visible in the chart, but it is ignored when determining which trend to select, as it cannot be considered a negative or positive emotion.

Perhaps you are interested in running a specific query on Twitter and analyse the results, please consult us.

Technical notes

The Twitter data is being extracted from Twitter and stored using TCAT (Twitter Capture and Analysis Tool). The subsequent emotion recognition is achieved in the standard fashion, by interfacing with our remote API from PHP and analysing tweet by tweet.

Copyright

Permission is granted to distribute the charts and information on this page, provided you link back to the original website.