FreshNetworks Blog

Spade stuck in the sand (Photo credit: Simon Cocks)

We’ve written before about the privacy issues that big data will raise. As we all become more aware of how the data we leave in our wake on social networks can be used, we will start to take more control over what we share with whom online. The updated personal Facebook analytics tool launched today by Wolfram Alpha is a great tool to understand what the data you share can reveal.

The original Wolfram Alpha Facebook analytics launched in September 2012 created a set of statistical insights into what you shared – how many links or photos you posted, at what time of day and where you checked it. It also included a cluster analysis of your friends – showing who was connected to whom.

The new Facebook analytics report moves beyond simple mapping of friends to begin to map out how your social networks is composed. Who are the key connectors, and who are the potential gateways to others you might not yet know. In total, Wolfram Alpha has identified five different roles in a social network:

• Social Insiders and Social Outsiders – two opposing groups, a Social Insider has lots of friends in common with you (such as a friend from university) whereas a Social Outsider has few friends in common (maybe somebody you met once on a holiday)
• Social Gateways and Social Neighbours – two more opposing groups, a Social Gateway has lots of friends who are outside your network (such as the editor of an online news source) and a Social Neighbour has few friends outside your network
• Social Connectors – the final group connects different parts of your network (so a university friend who you now also work with)

This analysis moves beyond grouping friends by reported facts about them (who you went to university with, who you live in the same town as etc) and starts to analyse how your network is composed.

This is exactly the kind of analysis that anybody who you give access to your Facebook data can gain – and indeed that any of your friends give access to depending on your own privacy settings. This kind of data is fascinating for us to look at, and is useful for many brands to understand and use.

Explore your Wolfram Alpha Facebook report

Like (Photo credit: afagen)

The more we contribute, communicate share and talk online the more we leave a trail of personal data in our tracks. This may be data about what we say to whom on Twitter, when we are most active or the photos we take. Or it may be data that we have captured from a specific activity – data on every run I have done in the last two years is stored by Runkeeper, for example. To have such a constantly growing, structured personal data set is very new and it offers real opportunities for brands and platforms. But also for individuals themselves.

The quantity and depth of data that we are structuring about our lives even on one network comes as a surprise to many people. Taking Facebook as an example – the data we create about ourselves and our networks is vast, and often hidden from the consumer – you just can’t imagine what it might be. The first step to help you understand the amount of data you have stored and how it might be useful is to visualise it – and search engine Wolfram Alpha have now produced a report that takes this information and presents it back to you.

For any user what you uncover about yourself, what Facebook knows about you, is interesting. For example, the word I have used most frequently on Facebook is ‘run’. The peak time for me to upload photos is apparently 9pm on a Saturday. And the most common first name and surname among my friends is ‘James’.

But what is more interesting to start to explore is how this Facebook data is able to understand data better than we might be able to. Take how it clusters my friends. Just looking at connections (and their connections) you can start to map out how my friends group themselves and really start to understand something about me.

Friend Network: Matt Rhodes

You can see three clear groups:

1. A tight cluster of yellow connections – people who are all interconnected and clearly all know each other. These are people I’ve been friends with since University.
2. A relatively tight cluster of blue connections – less interconnected but the groups of people I’ve made friends with in 10 years in London.
3. A more spread our cluster of green connections – a loosely connected set of people that I have worked with.

There are also the odd random connection that I have seemed to pick up along the way.

So Facebook can accurately and clearly summarise my friendships and how they interact. And you could probably make inferences from that about how likely I am to mix people across these groups – only a small number of people connect between the clusters, suggesting I am more likely to socialise in these groups separately (which to be honest I am).

There a lot of data out there, data that we are leaving in our wake with every social interaction. Currently this data is being used by the platforms and by brands, but the exciting opportunity is to see how individuals can take more ownership of their own data and get more value from it. The first step is to start to understand what data there is out there and how it is structured. The Wolfram Alpha Facebook reports make an important first step to revealing this.