Knowledge Matters™

It’s both frightening and mind-boggling to read about what is possible these days. Jure Leskovec from the Carnegie Mellon University, and Eric Horvitz from the Microsoft Redmond Research Laboratory, collaborated to examine 30 billion Microsoft Messenger conversations in a month between 240 million people! Their study appropriately titled “Planetary-Scale Views on an Instant-Messaging Network ” produced a network graph with 180 million nodes and 1.3 billion undirected edges, making it one of the largest network studies of its type. (240 million people had Microsoft Messenger accounts, but only 180 million people used them in the month data were collected.)

One of the key findings is that on a planetary scale the average path length between Microsoft Messenger users is 6.6 steps, and the median is seven. (The longest path was 29 steps). Theoretically this means I can get to anyone of the 180 million people represented as nodes in about seven steps. The authors extrapolate their data, and say they expect their findings would hold true if everyone on Earth could be included. Wow seven steps from everyone - now that’s connected! But is this actually the case?

Well here in Australia I know the Governor-General’s daughter, and the Governor-General knows the Queen, so for me that is three steps to the Queen. Now the Queen of course knows all sorts of highly influential people, including the Dalai Lama. The Dalai Lama probably knows the head lama of the Kyegu Monastery in Tibet, who no doubt knows a local monk or two. It’s probably reasonable to assume the local monk knows a least one person in Kyegu village, but can I get to talk to that person given they are seven steps away? Of course I can’t! I might be able to get two steps along the journey, but I am certain I can’t complete three yet alone seven steps. So what have Leskovec and Horvitz proved?

I think all they have demonstrated is the pervasiveness of social media, particularly in the 15 to 30 years age group. Take a look at Figure 10 on page 12, and the maps on pages 13 and 14, to see just how pervasive this technology is.

In my mind there are a number of ethical issues associated with this study. The authors say:

“All of our data was anonymized; we had no access to personally identifiable information. Also, we had no access to text of the messages exchanged or any other information that could be used to uniquely identify users. We focused on analyzing high-level characteristics and patterns that emerge from the collective dynamics of 240 million people, rather than the actions and characteristics of individuals”.

The sheer volume of traffic - 7 billion exchanged messages per day – probably means that even if they did have access content analysis would be almost impossible. That said I would love to see the ethics approval process they went through, because their statement implies it is possible to access the content of the messages, all of which presumably are private exchanges.

It’s also interesting they distinguish between the communication network with 180 million nodes and 1.3 billion undirected edges, and a “buddy network” with 240 million nodes and 9.1 billion edges. The buddy network was derived from contact lists – that is a link was made between two people if they appear on each other’s contact lists. It’s not clear to me how contact lists were obtained.

So to conclude it seems potentially we are seven steps and no more than 29 steps from everyone, especially if we use Microsoft Messenger. Big brother is watching and capable of collecting and analysing your most personal data on a scale hitherto unseen. Does this matter?

Regards Graham