Earlier this year a mass movement occurred in Bangladesh, which received little global news coverage. It was an immensely important event to Bangladeshi’s at home and abroad. This prompted me to try and illustrate the event with Twitter data myself, merely utilizing some free web services and a few hours time. Amazingly the results are beautiful and informative visualisations which did not cost a any money and very little effort and time.
Background: The Shahbag Protest
On the 5th of February 2013, unnoticed by most of the world, Abdul Quader Mollah, a Bangladeshi politician of the country’s marginal Islamist party, was sentenced to life in prison by Bangladesh’s International Crimes Tribunal, which investigates war crimes committed in the 1971 independence war of Bangladesh. The same day, online activists organized a mass movement in Shahbag square, Dhaka, Bangladesh, initially to demand capital punishment for Mollah.
The non-partisan protest rapidly gained public support and turned into a mix of remembrance of the 1971 war crime victims, and the peaceful youth movement in support of keeping Bangladesh secular. This is an extraordinary development in a country that has been polarized between two political parties, and a country used to regular hartals — organized violent political strikes. The Shahbag demonstration has been going on 24/7, and the recent killing of a blogger associated with the protest reignited the conviction of the demonstrators.
4 ways to visualize Shahbag on Twitter
I have been covering the Shahbag movement’s Twitter presence for The Daily Ittefaq and ClickIttefaq, a Bangladeshi newspaper and news site. I was amazed at how easy and inexpensive (read: free) it is to create stunning visualizations of what is happening on Twitter.
I employed four websites to dive deeper into the protest’s Twitter topic #shahbag.
The Twitter StreamGraph by Neoformix Java Applet queries for the most recent 1,000 tweets containing a search term, and visualizes a stream graph of the most common co-occurring terms. The user can then hover over the keywords and see the latest tweets. The example below shows a spike of tweets discussing the Bangladesh National Cricket team joining the protest.
Revisit by Moritz Stefaner visualizes tweets in real-time for a given topic. It illustrates tweets through the Twitter account’s avatar, and scales and interconnects them based on re-tweets and mentions. The user can highlight a single tweet and its impact by clicking on it.
The default overview beautifully and concisely illustrates a topic based on hundred(s) of tweets at a time.
Social Collider is a visual feast and exciting concept. You can search for a Twitter account or topic on the website and select a timeframe to visualize. The collider then arranges tweets vertically along a topic or user account. Tweets that relate to each other by content or directly are connected horizontally by a curved, or spiraled, colored line, inspired by particle collisions.
Besides its beauty, it illustrates strong interconnections in a dramatic, wild entanglement of lines. It also outlines global news outlets like BBC or CNN, and their relationships to the topic, by drawing their streams tangent to the core of the activity, since they reported about Shahbag sparsely.
Another interesting aspect of the online activity is its geographic distribution. Trendsmap provides a free (and paid) service to illustrate current topics on a map. It displays what is being discussed across the world, or reversely illustrates where in the world a topic is active.
The #shahbag topic appears only significant in Bangladesh, India, and Britain, the latter of which has a large Bangladeshi community. Visualizing the trending topics in Bangladesh confirmed the significance of the event with all major trending topics relating to Shahbag.
The quality of these free services impressed me deeply. None of them was a complete analytics solution that comprehensively analyzed the Twitter firehose, but each of them used a substantial sample and a unique interpretation, adding an important aspect to the whole story. Combining the single pieces shaped an insightful bigger picture of the topic.
Social media, big data, and visualizations are advancing rapidly, and I am excited to see where the journey goes.
This article was written by Christian Prokopp for and first published by the Big Data Republic.