Social graphs - scary and beautiful

If you are on Facebook you might have seen pictures from your friends where they post their "social graph"? These are indeed beautiful illustrations of your friend's network, and below I point to a few of the available "click and view" apps that will do this for you.

From all of these you will be able to quickly get an overview of your social network, and often you will be able to identify groups among your friends. As a bonus you will also find a tool that displays your social graph on LinkedIn.

But these are not only nice visualizations. They also invite to a moment of thought on how much a simple list of your friends, and how your friends know each other, can reveal. After presenting three visualization tools for Facebook I'll discuss a bit more about this. But first the tools:

myFnetwork
This visualization is the simplest of the tools. No interactivity, but you can zoom and pan around in the picture. Individual nodes are colour-coded based on some centrality measurement - the more yellow/white, the more connected to the surrounding nodes.

All nodes are placed in a circular patten, and you can identify some of the major clusters among your friends.

You can try myFnetwork here (link).

Social graph

This tool provides some interactivity, and includes the image of each friend in the visualization. The clusters of friends are more easily explorable, and some of the more obvious clusters are highlighted with pink circles.

You can hoover over a node to see the name of that specific friend. A click on a node will bring you to the corresponding Facebook profile. By clicking and holding down the mouse button you can grab a node and move it around to see which other nodes/friends that are connected.

Try Social graph here (link).

TouchGraph

The last Facebook tool to visualize you social graph is also the most advanced.

It does a nice job in colour-coding different clusters in your network and gives the nodes different sizes (which seems to be related to the number of friends in your network (degree)).

The graph is interactive, and can be configured and explored in a number of ways. Note that by default it also includes yourself as a node, but that can be changed in the advanced settings.

TouchGraph can be found here (link).

Some theory...
All of the graphs above are sometimes referred to as egocentric, level 1.5, social graphs. This means that the start form an ego (you) - but as you are obviously connected to all your friends it doesn't add any information to include yourself (rather, including yourself will in many cases make the network structure less obvious). 1.5 means that the network includes all of your friends (level 1), and how your friends are connected to each other (level 1.5). A level 2 would mean including all of your friend's friends, also those that not are connected to you. This has some further value, but on the other hand the Facebook API does not provide access to this information, and the number of nodes would quickly be very large.

It is quite obvious that by only looking at your friends, and how they know each other, you find some interesting patterns. Especially the groups (clusters) usually match very well to real life social groups.

The clusters are most often visualized by a combination of a layout algorithm, where groups will form, and by specific algorithms designed to calculate clusters.

With other tools you can take the analysis to the next level - there is a number of metrics that can be applied to these kind of graphs - but that is probably a topic for a future blog post.

...and the scary part!
Well, isn't the scary part clear by now?

Simply by looking at your friends, and how they are connected to each other, a lot can be seen. The relationship and context of a specific friend can be deducted in many cases. If you know one node that is in your family, you are likely to find the other family members in the same cluster.

Probably obvious to you, but as your social graph information is more or less openly available (on Facebook, but also on LinkedIn, Twitter and other networks) it is not that hard for anyone to have a closer look. And of course the network operator have a very nice picture of you - not only from what you "post and like", but especially from your social graph.

The bonus - a LinkedIn tool
Oh - yes - I promised you to also include a tool to visualize your social graph on LinkedIn. You can find the tool here (link). It is very similar to the other tools above, but instead looks at your LinkedIn social graph.

The main difference, besides the nodes and network as such, is that this tool include yourself as a node.

It does a good job in identifying clusters in your more professional network. (Note, you need at least 50 connections on LinkedIn to be able to use it).

Now it's your turn
Go ahead and try one or more of the social graph visualization tools I mention here. I'd love to hear your viewpoints, just leave a comment!

Also, if you know any other nice tool to visualize various social graphs I'd be interested.

Lunch, anyone?

It's not often I get excited enough over a new web site to jump into the editor and write a blog post immediately. This is however one such occasion.

In my feeds today I noticed several references to a new web site: YouGuysShouldLunch.com (or LunchaIhop.se in Swedish). Give it a try!

I haven't looked elsewhere, and there might very well be similar services around (if so, let me know in the comments).

However, there's several reasons that make me think this is simply brilliant, besides that the man behind (Jonas Larsson, @byBalsam) is from Sweden:

• The way it brings your online network of contacts into real world, physical meetings.
  
  
• You can't include yourself in a lunch proposal. You need to think seriously around who of your contacts you want to put together, and not simply propose a lunch with you and a friend.
  
  
• The use of LinkedIn indicates that "it's serious business" and provides both easy access to the professional background of a proposed lunch date. You are also likely to have some kind of professional relationship to the people you put together for a lunch date.
  
  
• It is a nice usage of the LinkedIn API. Mashups are great...
  
  
• The site has both a Swedish and an English version, with different domains. Nice detail.
  
  
• The logo indicates it can be a simple lunch, doesn't need to be at a fancy restaurant.
  
  
• ...and the list goes on with many small details.
  

However, the really neat thing is that this service is also one of the best and creative "I'm looking for a new job"-pitches I have seen. Ever.

As Jonas writes on his blog, he will leave his current assignment at the end of the year. One of the options when you pair together lunch dates is to invite one of your contacts to lunch with Jonas himself. He even provides a wish-list of people he would like to have lunch with.

No doubt he will have his calendar full of lunches the next months, and a bunch of interesting new job opportunities.

(Unless he builds a business around this initiative. Already valuable and interesting with several potential new things to add: How to find a good time for lunch (integrate with calendars), Where to lunch (find a good, close, spot by integrating with restaurant guides and maps).)

So - what are you waiting for? Which of your LinkedIn contacts do you think should have lunch together? (And who do you think I should have lunch with?)

The Pareto (80/20) rule and time management - the fallacy

The Pareto rule (or Pareto principle), also known as the 80/20 rule, is often used when discussing time management. However I do believe there's some flaws when applying the Pareto principle to time management, and I'll try to sort them out here.

The argument within time management often goes (after introducing the 80/20 rule):

"Thus, since 80% of you achievements comes from 20% of your work, you shall strive to focus more on you time and effort on those (current) 20%."

Right? Not!
At least I don't think so.

Don't get me wrong - time management is useful. But we shouldn't apply the Pareto rule here. Instead it all boils down to "do the important stuff first", but you should also be serendipitous.

There's a few potential and actual flaws that can be identified, but first - let's have a closer look at the Pareto principle.

The Pareto principle and power laws
The Pareto principle was named after Vilfredo Pareto, who observed that 80% of the land in Italy was owned by 20% of the population.

In a broader scope the Pareto principle is an effect of power law distributions, or more specifically - the 80/20 characteristics is a property of some classes of power law distributions.

Power laws are quite interesting, and well worth a deeper look (which I will spare you for the moment). For our purposes here we will note two things though:

• Phenomena described by power law distributions are scale-free, ie shows scale invariance. This means that the phenomena will always show a power law distribution, regardless of how you "zoom in" when you look.
• Power law distributions are an indication of a system in a critical state, or a system which shows self-organized criticality.

If you have looked into eg chaos theory, or read eg The Black Swan or Ubiquity, you will recognize that such systems are not predictable in the terms that we normally are used to with systems that shows a Gaussian random behavior. The changes are not distributed around an "average value". Instead the changes can be very large or very small, without any predictable size.

Now, let's have a look at the time management 80/20 fallacies...

False analogy (?)
The first objection - can you apply the knowledge from Italian land distribution to time management, or is this a false analogy?

To be honest; I'm not sure. I wouldn't be surprised if "how you spend your time" shows a power law distribution.

But I haven't seen any study proving this. Just because we "know that we spend a lot of time (80%?) on things that are not really productive (20%?)" it doesn't mean we have a Pareto distribution.

I would be happy if you can point me at studies that confirms that we do have a power law situation (or disproves...)!

But let's assume that we indeed have a Pareto distribution, and that the 80/20 rule do apply! What would this mean?

We will still have an 80/20 situation...
...even after we have done our "time management homework".

Remember that the Pareto distribution (and power laws in general) means that we have a scale-free behavior. Thus, after having cut of the "non-productive tail" of our efforts, the "productive head" (ie the 20% of our efforts that produces 80% of the value) will still have the 80/20 property!

What now? Should we iterate, and again cut of the tail (pony-tail?) of the head?

Also, how do we know that our initial 80/20 distribution isn't the result of a previous (unconscious) time management activity? Where should we start, and when should we end?

We don't know what will make up the "tail"
Another property of systems showing a Pareto/power law distribution is the unpredictability.

Yes, we should be able to analyze our past schedule, find the patterns and identify which part of our time was the unproductive tail.

However, if we do have a Pareto situation, it is not possible to foresee if any of the past "valuable time" eventually will be worthless, or if any of the potential "time wasters" will turn out to be the next big thing.

And it's all interconnected...
The last thing to note is that you activity that you'll trying to time manage consists of a number of smaller tasks. These tasks are not independent of each other, they all interrelate.

Going back to the theory, we can note that self-organized criticality is a property of some complex systems. "A complex system is a system composed of interconnected parts that as a whole exhibit one or more properties (behavior among the possible properties) not obvious from the properties of the individual parts"

Thus it would be impossible in most cases to filter out the less productive activity and foresee how the non-execution of that activity would affect your overall productivity, especially how it would affect your more productive time.

This might sound too theoretical, but it is easy to see the connection to eg the value of collaboration and networking.

• Perhaps those 15 unproductive minutes you spend with a colleague drinking coffee and chatting will result in that a piece of information you picked up in the chat saved you two hours of work later?
• Perhaps those 10 minutes you invested in helping a colleague solving a task not relevant for your performance will result in getting help later from that colleague, again saving you hours of work? (Not mentioning that those 10 minutes of your expertise you invested probably saved an hour of your companies overall time that your colleague would have spent on their own trying to find a solution).

The summary, and the common sense of time management
So, I don't think the 80/20 reasoning shall be applied to time management, at least not the way it is most commonly done today.

• First, please prove to me that we do indeed have a Pareto distribution here.

• Second, if we do have a Pareto distribution, whats' the implications of that?

But as I said in the beginning - time management is important, of course. But to me it all boils down to:

• Focus on the important things...
  And define what you mean by important. Most valuable? Most fun? Best investment for the future? Most urgent? Also, define "for whom". For you? For your stakeholder? (And who are they?) For the society?
  Tricky questions perhaps, but to me this are the important time management questions to have in mind.
• ...and be serendipitous
  Be open for new things, you'll never know when and how they will come in handy.
  Remember that it is often in the intersection of different fields or activities that the really valuable new thoughts and insights emerge.
  And you'll never know what the impact and future value of an activity that seems less valuable today might be tomorrow.

Where do all good ideas come from..?

...turns out that coffee *is* a good source, for two reasons...

I just discovered two videos with Steven Johnson (@stevenbjohnson) that are worth a few minutes. Also, he has recently published a book on the topic (Where Good Ideas Come From: The Natural History of Innovation), which I have added to my "to-read"-list.

First, here's an animated version of the story:


Next, here's a talk at TED by Steven Johnson. In here you will find out the two reasons coffee might be good for creating ideas (one was revealed in the video above).

Also, don't miss the story of how the embryo of GPS sprung out of the Sputnik launch in the 50's. This story is told at the end of the TED talk.

Four million Facebook users in Sweden - some statistics

Yesterday Facebook reached four million users in Sweden.As I have looked into the statistics and demographics at the one, two and three million mark I'm continuing the tradition with this post.

In Sweden 44% of the population are active on Facebook. It is especially interesting to note that more and more of the Facebook users are 40+ old.

In the age group 15-64 we see that 61% of the Swedes are active on Facebook. If you had any doubt, Facebooking has for sure become a mainstream activity.

The evolution over time
The graph below show the number of Swedish users on Facebook since the 25th of October 2007:The one-million mark was passed on October 31st, 2007. Almost 71 weeks later (March 8th, 2009) the two-million mark was reached. It took 41 more weeks to reach three million users (on Dec. 21st, 2009). The latest million of users, up to four million, took 36 weeks and was reached yesterday (Sep. 1st, 2010).

The gender distribution
Overall in Sweden 48.6% of the Swedish Facebook users are female, 45.1% are male and 6.3% have not defined their gender.

A further breakdown of the gender distribution per age group is in this graph:


The age distribution
The penetration per age is displayed in this chart:

The 100+ percent penetration between 14 and 23 years of age probably have a number of reasons, including:

• You have to be 13 to sign up for Facebook. If you are younger you'll probably fake your age. And once you have set your birth year you can't change it (you can change the date and month though...).
• It might also be that people maintain two or more Facebook accounts.

What do you think?

...and not only the youth on Facebook...
It's clear from the graph above that Facebook is used across the whole population. The penetration does not fall below 50% until the age of 42, and all the way up to the age of 60 the penetration is over 20%.

It is interesting to compare the penetration per age group today with previous occations. The chart below shows the penetration per age group at the stage where the number of Swedish Facebook users passed 1, 2, 3 and 4 million users:

Another way to view the data is to look at the accumulated penetration as we go from the younger to the elderly:

Initially (at one million users, the blue line) 80% of the users where younger than 30 years. Now (4 million users, green line) we find that we have to move to the 40-44 years segment to cover 80 percent of the users.

That's it for the 4 million report. Now let's see if (or when..?) Sweden reaches 5 million users on Facebook.

Any thoughts? Will the growth continue? Or will another site start to replace Facebook as "the place to hang out".

(Note 1: the number of users is measured through the Facebook advertising tool. Although it is not clearly stated what criteria is used to classify users, the best guess is that A) the tool measures active users, which probably means users that have logged in the last month and B) The nationality is to a large degree decided by looking at the IP-address range of the user.

Note 2: The data on the age structure of the Swedish population I used to calculate the penetration is from SCB. However, the latest data on the age structure of Sweden available as of today is the structure for 31st of December 2009. After some consideration I took the liberty to use that data, but to add 1 to all ages. Clearly this is only an approximation of the current age structure, but as eight month has passed since Dec. 31st I believe that is to some extent a better approximation.)

Gothenburg and Facebook - what's up?

I just had a quick look at the regional/local statistics for Facebook in Sweden. To my surprise Gothenburg seems really underrepresented.And yes - I also used Gothenburg (and even Goteburg) as search terms - all gave a similar figure.

As a comparison - the same search for Stockholm returns more than 2.6 million people (far more than the official population in the region?) and for Malmö it returns 234 000 people.

The official population of Gothenburg is just above 500 000 individuals (Gothenburg is the 2nd largest city in Sweden).

How come? Any ideas? Could it be that:

• People in Gothenburg doesn't use Facebook and social networking sites to a large extent?
  
  
• Perhaps there's another social networking site which is large in Gothenburg?
  
  
• People in Gothenburg would rather not state that they live in Gothenburg?
  
  
• A slightly more technical possibility - Facebook uses, among other things, the user's IP address to decide the geographical location. Perhaps the Internet connections in Gothenburg are routed through Stockholm or some other place?

I'm stumped and surprised. What do you say? Do you have any Facebook friends from Gothenburg to ask..?

Copyleft or copy-theft?

Creative Commons and Noncommercial use - what does it really mean? Can you help me understand this?

Please take the time to answer this poll (link) - I will post a summary when I get sufficient number of replies.

You might be familiar with Creative Commons and how they:

provides free tools that let authors, scientists, artists, and educators easily
mark their creative work with the freedoms they want it to carry. You can
use CC to change your copyright terms from "All Rights Reserved" to "Some
Rights Reserved."

Creative Commons works with four basic attributes, which are combined into a specific license:

Now, it's the "Noncommercial" I need your help with. (Again, here's the poll where I would love your feedback!)

How shall "Noncommercial" vs. "Commercial" purposes be interpreted? Is there a clear definition?

• It is pretty clear that you can't sell somewith which is licensed as "Noncommercial". Ie, you can not make a postcard or a T-shirt of a non-commercial picture and sell it.
  
• On the other hand I guess you can make a T-shirt and wear it yourself, or make a postcard and send to you friends?
  
• Can you make a postcard or T-shirt and give it away? To your personal friends? On behalf of your employeer? If a corporation gives something away, even for free, it is in many cases a commercial reason for this.

It is quite common to apply a copyright model that let people reuse and embed content from the Internet (eg video files) given certain conditions. In many cases this includes a condition that the content "may only be used for non-commercial purposes".

At a first glance the interpretation of this sounds obvious. But beware, the definition of "Non Commercial" might not be as obvious as you thought. The interpretation is not crystal clear, and the formal meaning is in many cases more narrow than you might intend - both as a content creator, granting a "Non Commercial" license and as publisher/user, using "Non Commercial" content.

Let's have a look at the topic, starting with Creative Commons and then moving to some examples.

Creative Commons
Have a look at Creative Commons and the set of licenses they offer for creators and contributors that are looking for a crisp and clear way to apply a copyleft approach (rather than copyright) to their works and creations. In essence, this allows creators to apply a "some rights reserved" license to their works.

One of the attributes that can be part of a Creative Commons license attribution is the Non Commercial (NC) tag. Three out of the six available CC licenses includes this tag.

Now, what is meant by the NC tag? It turns out that this is not as obvious and straight forward as you would wish.

The license attribute
The short version of the Noncommercial attribute reads:

You let others copy, distribute, display, and perform your work — and derivative works based upon it — but for noncommercial purposes only.

When you dig into the small print of the more legal version of a CC license including the NC attribute you find the text:

You may not exercise any of the rights granted to You [in Section 3 above] in any manner that is primarily intended for or directed toward commercial advantage or private monetary compensation.

What does this mean?

The key phrase is "primarily intended for or directed toward commercial advantage or private monetary compensation".

This phrase is not as clear as you would wish, and in many cases I believe the practical interpretation applied by individuals differs. Especially as many probably doesn't read the fine print, and instead relies on the even more open-ended phrasing "noncommercial purposes only" in the Noncommercial attribute.

The problematic word seems to boil down to "primarily".

Primarily or Secondarily?

How do you define what's primarily and what's secondarily..?


Consider a few examples:


Private use:

• You have a blog where you post "funny things" just for the fun of it.
  
• You have a blog where you post "funny things" just for the fun of it. Also, just for the fun of it, you included some ads on the blog.
  You know that you'll probably not make any money from the ads (best case a few dollars a year). The reason you included ads are not to make money, but something else (like "I'd like cool ads on my blog" or "I'm trying to understand how ads work technically").
  
• You have a blog where you post "funny things" just for the fun of it. The blog is hosted at a site that includes ads automatically (like eg. Facebook).
  The reason the ads are there is to finance the (free) service that hosts your blog.
  
• You have a blog where you post "funny and useful things". The aim is to make yourself more attractive to employ.

Corporate use:

• You are posting to a corporate, external blog. The aim of the blog is to promote the "know-how" of the company and to attract new customers.
  
• You are posting to a corporate, external blog. The aim of the blog is to build the perception and brand of the company. All sales goes through other channels.
  
• You are posting to a blog on your company's intranet. The aim is to share knowledge, which in the end might make your company more efficient and competitive.
  
• You are posting to a corporate, external blog. The company is a non-profit organization, but you still charge for your services in order to finance the activities you do to eg. support victims of various disasters. The aim of the blog is to promote the "know-how" of the company and to attract new customers.

In which of the above cases would it be OK to use content with the Non-commercial attribution? Have your say in this poll.

A study from Creative Commons
Creative Commons published a report September 2009 on this topic, looking into how the online community defines and perceives the term Noncommercial. Interesting and recommended reading - but no clear conclusion of the definition. Read the announcement here.

Conclusions?
So, what's your take on the definition of Noncommercial? A no-brainer or a can of worms?

How do you apply the definition, or don't you you see this as applicable to you?

Please drop a comment below, and contribute to the poll!