Data Mining: Text Mining, Visualization and Social Media

Klostu: 300 Million Users in a Single Social Network

Ron Kass, CEO of Klostu, has given me a tour of the system and I now have a pretty good ideal of the complete package. The big picture is as follows. There are 300 million message board authors out there. In general, each message board requires an individual identity (author name, profile, password, etc.). In addition, none of these boards exist in a common, well defined infrastructure. Klostu aims to solve these issues.

Klostu offers a single web identity - like a board or blog profile with no associated publication. Your Klostu home page (here's mine) is composed of the now familiar drag-and-drop editable AJAX style page. Note that there is nothing here that prevents the addition of personal content - like a blog - in the future.

Klostu has formed relationships with a number of message boards (and plans to grow this number rapidly in the future). Participating boards install some active elements in their pages so that you can log in with your Klostu identity on that board. This then brings along your profile as well as a number of other standard board elements such as avatars and (in the future) signatures. Consequently I can create a single board identity and use it on multiple boards.

In effect, Klostu aims to position itself as the interface to the largest group of online identities (one might even call it a social network). If things take off, there are many potential benefits that will come to authors via a single identity. Note that there certainly will be cases were individuals want to maintain different handles in different domains.

Naturally, one also has to think through the monetization of this system. One obvious are which Ron discussed with me is that of advertising. With relationships with many message boards on the one hand and landing pages for author identities on the other, they would seem well positioned to funnel some form of advertising network content.

It is early days just now. I'm hoping that Klostu will start releasing some key stats soon (like number of users and number of boards registered).

Klostu - Wraps Off

Klostu - the new service from the company that brought you BoardTracker - has taken of some of its stealth wrapping. What is revealed is somewhat tantalizing. Klostu is a portal into the boardscape, it provides a single surface from which one can access different boards and encounter board authors. It also provides a single identity - a Klostu ID - that can be used (somehow) across all boards. There is more information available on the about page which also mentions:

While Klostu offers all the social networking and ajax desktop interface features expected today (and more), a key feature which emphasizes the real power of klostu is its function as a platform to bring other web services to the boards. Through klostu, a whole range of 3rd party products can be fully integrated within boards thereby easily extending their reach to millions of people active on the boardscape.

I'm really excited to see this level of innovation coming from a company that has a clear passion for social media and a vision which encompasses the larger, but less trendy, boardscape. Congrats to Ron and the rest of the Klostu team.

Mapping The Boardscape

You've seen plenty of graphs on this blog describing the blogosphere. What I'm showing below is an initial application of the system I use to generate those maps applied to some data from a message board. This particular board discusses the Sony Playstation (I'll post more about it later). I actually think that this is a slightly more interesting graph in terms of the social network. (Boardscape is a term coined by Ron Kass of Boardtracker).

Note that the data here is just a sample - I'll look at getting a more comprehensive set sometime soon.

BoardTracker Advertisement

I just spotted an advert for BoardTracker (one of the best message board search engines out there) over on BattelleMedia. I don't think I've ever seen an ad for a social media search company on a blog before. Perhaps BoardTracker is gearing up for a jump to the next level...

BoardTracker Update

Some updates on BoardTracker's stats.

Forums over time:

Threads over time:

It looks as if a significant number of threads are strating to pour in.

BoardTracker Update

Readers of this blog will be accustomed to my updates on the message board space, especially BoardTracker, and my conviction that there is far more data out there on boards than there is on blogs. They will also be aware of my belief that the blogosphere is following main stream media in an ironic development which sees the A-List bloggers getting delayed on their observations, insights and scoops.

I'm really excited to see that Rubel is finally giving some pixels to the message board space:

Finally there's now a way for all of us to search message boards, not just professionals who license the high-end tools. BoardTracker.com is tracking 16 million conversation threads in 26,000+ forums. They aggregate content into tag clouds too.

While these numbers are very small, I still think it's smart for these guys to be focusing on forums. There's plenty of value here and they're filling a void in the market. Take health, for example. Most of the conversation around health takes place in forums, not blogs - at least right now. I would love to see more innovation in this area of conversation monitoring. Better yet, they should find a way to aggregate blog comments too. Wither not the forum.

I'd love to see his data supporting the statement that health issues are more the domain of message boards than blogs (I'm not doubting it, I'd just love to know where this comes from). Steve - what message board analysis tools have you been using to date?

Steve's suggestion that forum content systems should aggregate blog comments is, however, ill-conceived. Blog comments need to be tightly integrated with blog acquisition systems to make sense of the thread of discussion.

Tags in Board Space

A post on a BoardTracker forum pointed to the use of tags on Slashdot. In addition, the post highlighted the use of tags on BoardTracker's own forums. Sounds like there will be more tag based features to come:

Tagging is very popular these days, particularly among bloggers, and of course BoardTracker has brought tagging to forum threads too as you can see in the 'topic tags' above. For an example of these tags displayed on boardtracker have a look here..

http://www.boardtracker.com/search/?search=site%3Aboardtracker.com

We are rolling out several new tag related features for boards soon also.. watch this space! 

BoardTracker gets Trendy

BoardTracker, the message board search engine, is now displaying time series of the number of boards and forums that it tracks. I've been monitoring these numbers for a while and it is interesting to compare different views. Here is BoardTrackers graphs of forums and threads over time:

The data I've been accumulating is over a wider time window:

and

My graph for forums suggests that BoardTracker updates its index of forums on a weekly schedule.

BTW, an additional stat I'd like to see is the number of actual boards that are crawled (not just the number of forums).

BoardTracker Update

I've been monitoring the statistics published on BoardTracker for a few days now. My hope had been to see the number of forums and the number of threads showing a steady, monotonic increase. However, as the graphs below demonstrate, something quite different appears to be happening:

The second graph, showing the number of forums tracked over time, is probably the key here. The fact that it appears to flip between certain values suggests not that there is a gradual purging of, for example, dead boards, but that there is some issue with BoardTracker's infrastructure.

What is interesting also is that during the period at the centre of the second graph during which there appear to be not additions to or removals from the database of boards tracked, the number of forums shows a steady increase: the natural growth of the forums on a closed set of boards.

What's up?