« Passive Jelly Browser » : différence entre les versions

De Assothink Wiki
Aller à la navigation Aller à la recherche
Contenu ajouté Contenu supprimé
Aucun résumé des modifications
Aucun résumé des modifications
Ligne 17 : Ligne 17 :
:the '''passive jelly''' containing:
:the '''passive jelly''' containing:
::units
::units
:::concepts (language independant)
:::concepts (language independant, unnamable but definable objects)
:::percepts (language linked, lemmas, close to concepts)
:::percepts (language linked, lemmas, close to concepts and linked to them)
:::variants (language lined, flections, linked to percepts)
:::variants (language lined, flections, linked to percepts)
::links between units
::links between units

Version du 23 novembre 2011 à 16:01

Assothink is: 

  • a new model of Associative Artificail Intelligence
  • a set of programs implementing this model

Among the accessible programs are:

Assothink model overview

Assothink is an attempt to create associative (thus non-numeric) artificial intelligence. It is assumed that natural intelligences do not use computing devices: they perform various 'intelligent task' through simpler association processes.

The assothink model manages:

the passive jelly containing:
units
concepts (language independant, unnamable but definable objects)
percepts (language linked, lemmas, close to concepts and linked to them)
variants (language lined, flections, linked to percepts)
links between units
qualified links (structured links, the connecting object being a concept)
fuzzy links (with variable permeability)
the active jelly containing
excitation levels for all concepts, involving
excitation transition effects
excitation resonance effects
focused consciences organized in tree

Understanding the information displayed

The information displayed in the browser window is organized in bubble units and links.

Bubbles are used to show:

  • concepts
  • percepts
  • variants

Links (qualified and fuzzy) are used to connect the bubble units.

Moving the mouse on bubble units or links provide more detailed information of the focused objects.

User actions

When using the browser the user will explore detailed information by moving the mouse onto objects.

He can also:

  • click on an object, to make it the new center of exploration.
  • enter a word in the (lower left) text area (english or french). If no word is found, closest suggested words will be displayed.
  • click on the (lower right) backward / forward buttons for history navigation
  • click on the (lower right) random button to explore a random concept
  • click on the "Doc & Help" button to reach this wiki page 
  • enter .XXX where XXX is the reference