Subjecting Esoteric Practice To The Relational Machine

Over the weekend I held the first workshop of Divining Data – Mining (or Data Mining Divination, presently in two minds RE final title) at Galerie8’s launch weekend, set up alongside Flee Immediately’s stall during the A/FFAIR of Publishing Expo.

The code sprung a last minute debug on me (Perl and I just don’t get along sometimes) but ran seamlessly for the duration of the weekend, and for the most part my rough n’ ready Tk Interface did the job.

The experience unfolds as follows. The participant is seated  before a typical keyboard, mouse and monitor ensemble but the desktop furniture is joined by the modified tea cup I constructed. At the bottom of this tea cup is wide angled webcam which records the tea as the user imbibes it. The webcam logs data which is entered into the ‘computer centric’ tables of my tea reading database.

Once the tea is finished and the correct protocol followed RE channeling your unconscious into the tea cup (facilitated by three swift swirls in the users inferior hand) the user must then complete a bog standard data entry form. This form demands the user answer questions superfluous to the fortunes imparted by the symbols witnessed but useful in buttressing any later computer vision (CV) analysis of the tea leaves.

I was interested to realise that by loading this human-computation questions near the beginning of the data entry form that I was likely favouring the analytic element of the users pysche rather than the associative element most likely to be used in seeing symbols and shapes in the Rorshach like blobs left behind by the tea leaves. This is something I will change for subsequent versions.

In spite of this very few of the 15 participants had difficulty recognising shapes and to my delight the computer used it’s synonym look up table to ensure that symbols not originally in the dictionary provided by ‘Telling Fortunes By Tea Leaves’. I was also delighted to see that everyone enjoyed the process and some were pleasantly suprised by the spooky soothsaying accuracy of some of the Fortunes returned. Often this uncanny element was facilited when the symbols were read in accordance with initials witnessed in the cup, something which I explained to participants as I am not sure of an elegant way to code the computer to maximise this side of the fortune telling.

Shapes noted by the participant are logged in a database which then pattern matches these noticed patterns against a historical dictionary of what fortune correlates to what symbol. An overhead shot of the tea leaf pattern is also captured.

There are some kinks to iron out of the code, and the end user feedback needs further thought, but it is certainly progressing nicely and moreover the process of gathering data around an intuitive practice has begun. Thinking through how a data entry form can disrupt that intuitive form of knowledge acquisition remains the issue I most wish to address while the videos captured of the loose leaf tea being drank remain the most artistically compelling facet of the project for me.

Expect more prototype and code construction posts to follow over the coming two months. The next place I shall bring Data Mining Divination will be AND Festival in Liverpool

~ by Stephen Fortune on July 27, 2011.

One Response to “Subjecting Esoteric Practice To The Relational Machine”

  1. Hi.
    Aside from the time in Liverpool, when are you planning to do your next data mining divination? I would really like to try it.

    P.S. The sound installation began not long after you left on Sunday. Shame it didn’t start earlier – it was quite phenomenal

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: