Numbers with a view
How to give numbers a picture? A group of data collectors, data users and analysts in Bangalore has an answer. Part of Open Data, they are at work “visualising” data on politics ahead of the coming general elections.
Last month, at an Open Data “unconference”, a set of 32 slides termed ‘Visualising Politics’ was unveiled and put online (http://www.slideshare.net/gramener/visualising-politics) as an example of how “seeing” can make a difference in understanding cluttered tables that sometimes only the trained eye can make sense of.
Sensing the interest in the subject with crucial polls lined up, this group comprising professionals from the private sector as well as those associated with the National Informatics Centre and Census believes the data in the form they have presented can help voters, and those looking for connections between policies and how people vote, find some answers.
According to the group that first connected online, this is of particular importance in India given the sheer volume of data generated in the country, by the government, the RTI applications and other individuals.
The Bangalore ‘unconference’ was Open Data’s second annual meeting, and the key mover was Chief Data Scientist S Anand of Gramener. A Bangalore-based company, Gramener has been grappling with how to make “big” data in a country like India more accessible and readable, thereby allowing it to be used more intelligently and widely.
Says Anand: “We want to make slides that play with numbers and tables available anyway. But helping others visualise or see this as pictures makes it easy to assimilate and use the data.”
Plotting data for largely the 2004 general elections, the group has arrived at some interesting conclusions. For example, the percentage of votes polled is consistently inverse to the number of contestants in a race; and also, that the more densely populated a constituency, lower the voting. Another interesting data shows that as the number of contestants increases, the percentage margin by which a winner wins increases.
If some of the group’s data shows differences in voter participation across regions, another brings out that Udaipur was the only constituency in the country where women contestants outnumbered men in the 2004 general elections; the men lost their deposits.
The group is now working on other data, such as foreign currency received by NGOs, Karnataka Assembly results from last time and Lok Sabha attendance, mapped party wise.
Analysts familiar with government data collection have always maintained that India’s ability to collect data for various things, across vast and inaccessible regions and people, is commendable. However, even in large exercises such as the Census, the data is collected but without necessarily seeking answers to any correlations between variables. So, sifting in the data, there is a tremendous scope to look for specific answers to specific questions, which “visualising” helps make clear.
Several companies and NGOs working with Gramener use the data to plan the relationship between problems and policies better.
Amrtha Kasturi Rangan, who works with Arghym, a foundation working on rural sanitation, says “visualisation” of big data sets often seen as clunky otherwise is very helpful.
“The government set aside around Rs 3,500 crore last year (before revision) for rural sanitation. The data on the spends and achievements is available on their website. Our effort is to help translate these numbers into a more understandable format and we see the utility of having visualisations for this. This will help both decision makers and civil society understand progress of rural sanitation at a glance. We hope that these can be then used as a governance tool to help decipher patterns, trends, good examples and anomalies,” says Rangan.