Archive

Google has published their mobile and social network platform plans during the last week. Not so many surprises.

OpenSocial is quite natural next move from a company that wants to collect a lot of data. Social networking platform is a valuable source of behavior and social network data. And it is much more than have one or two social sites. The main concept with mobile has similar objectives. Who will dominate mobile usage and social data in the future. And finally who will prevail in digital media marketing.

But is it so simple that one company can prevail all relevant data in the future. Is it even possible that one company can collect all data when there will be more and more data all the time everywhere? I believe they can collect a lot of data. But on the other hand I believe that the services and data will be much more de-centralized in the future.

Intelligence comes nearer users and user devices. Smart analytics, customer profiling, and social intelligence come also to terminals (PC or mobile). The best solution to challenge Google is not to try to collect more data, but to have more intelligence to utilize the data. And this kind of analytics and advertising solutions help media and mobile companies to challenge Google.

Here at Xtract we get to work with data sets which are often in the form of networks. These may be, for example, social network data sets in online forums or links between web pages.

Often it would be useful to draw pictures of these networks, to understand better what their structure is. However, with networks that have millions of nodes, this is in practice impossible.

The problem of visualizing large networks is a common one and many different mathematical and aesthetical approaches have been taken to tackle it. Visual complexity is an interesting website, where a large number of network visualizations are presented.

When Arlinda asked me to write something about my master’s thesis on the company blog, I was stumbled. Thus is not because I wouldn’t have anything to tell about the thesis. Instead, I have been basically living the thesis for the last eight months, which makes it difficult to pick any single viewpoint to the work.

The topic of my thesis work is Bayesian clustering of huge friendship networks. It discusses methods to find structure in large networks, such as networks of friendships and presents an algorithm for this purpose, originally devised by Janne Sinkkonen, the chief researcher here at Xtract.

The algorithm finds overlapping group (i.e., cluster) structure of nodes. In each group the nodes are expected to possess similar traits and each node may belong to multiple groups.

The method studied in the thesis uses a Bayesian model of friendship formation, which is based on the theory of homophily studied by sociologists since 1950’s. What is nice about the approach is that it not only tells into which groups each node belongs, but also about the certainty of the group assignments.

It was interesting to do the thesis for a company, because there was a real need for the algorithm: The algorithm and its implementation are now being taken into use in customer projects. This is different from the academics, where the developer may often be the only one to ever use a reference implementation.

You can read more about the algorithm from the abstract we wrote for the MLG’07 -workshop: “Inferring vertex properties from topology in large networks”.