Alex II

The Model

History

Alex I was developped between November 2009 and february 2010.

Alex I was developped without any evaluation of other similar studies or research in the scientific world.

It showed some promising perspectives, but also some significant limits:

• the fusion beween words and concepts, which results in weaknesses, incoherences, conflicts, and unreachable areas
• the lack of a correct implication of numeric cognition and processing
• a too rigid description of the conscience

In februray 2010, the Alex II project started on various fronts:

• first contacts with Iridia (H. Bersine) and Cental (C. Fairon)
• various writings in French, thoughts and explorations rather than solutions and choices
• a model overview in English (anchored here), a full description of the selected solution
• the development of java classes forming a new Alex server side.

Overview

The following paragraphs and the linked pages describe the Alex II model.

Globally Alex II is organized in 2 levels:

• the jelly level, autonomous, passive, the raw material
• the conscience level, transient, dynamic, active, interacting, built on top of the jelly

Slowly dynamic part : the base jelly

The jelly includes nodes and links.

Basically, a node is a content holder. But in most case, this content is impossible to name and very difficult to describe!

Basically, a link is a connection between 2 or 3 nodes.

Links and nodes

Fives levels of nodes: concepts / words / variants / signals

Link classification

Concepts classification

Highly dynamic part : conscience and focus

A conscience is a tree of focuses.

The root focus is the only one to persist. Other are transitional.

A focus may be described as a thread (in java language understanding).

Focus object may be either target oriented or input-driven.