quit! Chain-Saw-Processor back!

Help! What the f*** is this thing?

Quick description and rationale

The Chain-Saw-Processor is inspired by artificial intelligence researches and various psychological theories such as :

Visitors are invited to explore the interplay of about 300 images and 100 key-words, and to build chains of thought with them, by being more or less constrained by these theoretical principles.


All right, but how does it work then?

Main page & processing

The main page is of course the index page, where the main image is displayed.
The PROCESS key allows you to ask the processor to change that image. This processing function compares the properties of the initial image with properties the other images of the database.
The next image is chosen as it matches some properties of the initial images. When the next image is then displayed, the link between this image and the previous is represented by the underlined word (and the previous image is displayed at the bottom of the page, in smaller size). And so on...

The EXAMPLE 1 below shows explicitly how images are linked by this process with matching keywords. The keywords' disposition and selection are described in more detail in the next sections.

EXAMPLE 1. Connection principle



More about tags...

Images are tagged on multiple dimensions, so it can be disappointing at first glance. However, the structure of the display is quite simple :

The EXAMPLE 2 below shows various keywords of one image and their specific display area, as listed above.

EXAMPLE 2. Image's tags



The Settings page

The image selection process is only partially random, and the SETTINGS (or control) page is designed to impose some constraints to this selection process.
The quadrant proposed in this settings page allows you to choose a degree of constraint in a 2-dimensional associative space :

Back on the main page, the selected degree is represented by one or two white key-word(s), whereas the others are grey or black (see example below).

Note that with default settings (as in the PROCESS' example above), no specific degree is selected (every types of connectors are equally probable).

The EXAMPLE 3 shows how the selection process is constrained, limited to certain type of keywords (connectors), and how the display in the main page will change as a function of these settings.

EXAMPLE 3. Settings' impact on selection constraints and display



The Chain-Seen page

Last but not least, the CHAIN-SEEN (or memory) page helps you to better apprehend the chain of words and images that you have processed. This chain is made of the images you have seen, linked by the word that has connected them during the process.

The EXAMPLE 4 below represent a possible Chain-Seen, corresponding here to images and underlined words (connectors) presented in example 1.

EXAMPLE 4. A possible Chain-Seen