We wanted to create an exciting way to visualise the social data generated from social data around Game of Thrones. We used data from twitter gathered with their Streaming APIs and processed with the real time computation library Storm to display a 3d visualisation using Unity3D to display to a user using an Oculus Rift.
To efficiently gather conversations about GoT we built an application that could access in real time to Twitter’s global stream of tweet data. To process and filter the large amount of data we used the Storm project library which transforms a stream of tweets into a stream of trending topics. Combined with a machine learning toolkit for the processing of natural language text we were able to filter top GoT related topics.
To identify trending topics we used multiple metrics:
Time: to be considered as trending a topic had to be mentioned in conversations not older than 15 minutes.
Popularity: a score was calculated on author popularity based on followers, number of retweets, number of favorites and conversation size.
We represented each of the trending topics on twitter using 3d cities. The amount of buildings and their height are used as a representation of their relative popularity compared to each other. Larger and taller cities represent more mentions of that topic.
Using the oculus rift we wanted to take the user on a flight over each of these cities similar to the GoT intro, seeing each of the cities grow out of the ground as the user passes over them. The user is moved over the cities on rails (along a set path) with the ability to look up, down and around at the cities and landscape which is procedurally generated around them.
Using familiar elements from the tv show we believe we have made an engaging way of understanding data.