The University of Reading’s workshop on “Big Social Data: Interdisciplinary Analytics”

The University of Reading’s workshop on “Big Social Data: Interdisciplinary Analytics” was held last week, with funding contributions from the University’s RETF and the Dept. of English and Applied Linguistics (DELAL), Henley Business School (HBS), School of Politics, Economics and International Relations (SPEIR) and School of Systems Engineering (SSE). The workshop was organised by academics […]

Democracy is Cyber-participation: General Election Twitter Boom

Thursday the 26th of March 2015 was the day of the TV debate ” Cameron & Miliband: The Battle for Number 10″. Arguably this has been the most remarkable day for Twitter activity related to UK politics so far. Many media have reported that allegedly as many as 260,000 tweets were generated during the event.

We have been using the Twitter streaming API to monitor any tweet related to UK politics in real time: a combination of more than 30 tracked terms and ad-hoc filters for a political context check are used to identify the ‘political’ tweets. According to the Twitter streaming API reports, we have missed only 5% of the total traffic generated by citizens who were inspired by the event to become “cyber-chatterboxes“. During the day of the event (26/03), a total of 388,733 ‘political’ tweets were collected, with a 300% increase w.r.t. the previous day. Only in the evening, between 20:30 and 24:00, there were 348,993 ‘political’ tweets, of which 119,282 specifically included the term “BattleForNumber10”. This chart reports the number of recorded tweets over time with specific counts for those that included an explicit reference to a political party.


This second chart provides a more detailed view of the frequency of party terms in the tweets. It clearly shows the transition Tories-Labour corresponding to the transition Cameron-Miliband during the event.



Welcome to the Reading General Election Blog, created as part of

We are a set of staff at the university with an interest in both big datasets and social media data:

Together we’ve formed the University of Reading Big Social Data Research Group.

We will begin to blog based on the election using social media data, but in the interim we will begin presenting data-informed comment based on our own research on here – stay tuned…