Thursday, March 11, 2010

Statistics for Humintell Products

Humintell has an appealing product line to train in recognition of emotions, microexpressions, subtle expressions and so forth. Their products are great training tools in general but some features are missing in my opinion. Statistics is a phenomenal tool to improve oneself to become a true master and it would be an invaluable addition to the existing features in Humintell's product line.

I am a passionate person and always go the extra mile. I have thousands of pages written with ideas and concepts but being a software designer also means to bring them to life. I enjoy taking up challenges, roll up my sleeves and get down to the deed.

Most web browsers allow user scripts written in JavaScript to be executed once a page is rendered. The user script as seen on the screenshot has been written using Mozilla Firefox, Firebug, GreaseMonkey, and vim.

My prototype retrieves and/or collects data on the fly that is stored in Firefox preferences. The persistent data is loaded each time you log on Humintell's training site and is used to render statistics. The practices and post-tests are averaged to have a better idea of the trainee's actual performances.

More than half of the population are predominantly visual. For this reason, visualization takes an important place in our lives. It often translates as charts and graphs when representing statistics, whether it is in a research report or a PowerPoint presentation. Most people who are not familiar with advanced statistics are drawn to use histograms or pie charts. They are all right but not necessarily the most representative depending on what the data is. I have instead chosen to use a radar graph since it allows to display multivariate data in two-dimension, it gives a unique perspective of the overall data and the relational rapport of each data with each other.

I will eventually publish my user script provided that Humintell is all right with the idea. It's perfectly legal to write (and distribute!) such user scripts that are ran on the client-side but it's important to keep a good relationship with companies around. Moreover, they could themselves implement the feature on server-side, which would be much more reliable and could use pChart to render the radar graph.


  1. Thanks for sharing! Wasn't aware of Humintell and will feature in my list! If you are looking for a good overview of available statistical resources worldwide: I have put one of the most comprehensive link lists for hundreds of thousands of statistical sources and indicators (economics, demographics, health etc.) on my blog: Statistics Reference List (almost 700 sites by now and growing). And what I find most fascinating is how such data can be visualised nowadays with the graphical computing power of modern PCs, as in many of the dozens of examples in these Data Visualisation References. If you miss anything that I might be able to find for you or if you yourself want to share a resource, please leave a comment.

  2. Hello CrisisMaven!

    I am not sure that you understand my article. This feature is of my own doing. It is not yet and may not be integrated in Humintell's products. Choice is theirs. Their products deal with emotions rather than statistics.

    You were perhaps looking for an opportunity to plug your blog? Well, that's all right. I have visited your website and it is delightful. A comprehensive list is exactly what it is.

    Keep up with your good work!

    Viele Grüße,

  3. Hello Ian, thanks for your kind words. No, I didn't misunderstand your article but was fascinated by it as I recently got hooked on the possibilities of visualisation - and yes I then also took the opportunity to "plug" as you call it. While I began my list purely from a statistical and indicator angle I discovered so many tools by now that allow to correlate and present data visually that I then decided to promote these ideas more stringently than is common as I believe it has great merits in education and I even have a feeling had these bankers looked at correlation graphically the crisis wouldn't be half as bad. And above all there's an aesthetic quality in it which I also saw in your work.

  4. I have been increasingly using data visualization in my projects the last 5 years or so. I will certainly post more articles in which I use data visualization. Stick around. :)

    By the way, will you cover other topics than economics? There are so many fantastic things done using data visualization in other fields.

    Do you have anything to represent the production of electricity in a power plant (eco-energy or whatnot)?