Creative architectures for problems solutions

Based on the exposed, it becomes relatively simple to model a "model of creativity" seeking to make possible its implementation in artificial intelligence system. Pict. 3 presents a general vision of this architecture:

Picture 3: Model of Creativity

This architecture offers a linear visualization of the creative process regarding the solution of no-trivial problems. The visualization is offered this way to facilitate the didactic apprehension of the model. In the real life, the mental processes happen in parallel, in several areas of the brain. They were classified in three categories: the everyday problems, corresponding to the situations of daily life (how to change a lamp? how to tie the shoes?); the difficult problems, which use the mental powers of convergent thought (Guilford, 1950) and find solutions by logical-deductive processes; and the complex problems, which require the creative capacity for their resolution.

The "Domain" box represents the abilities and the individual's competences, according to the classification of Czikszentmihalyi (1988). In the initial evaluation and categorization of a problem, the domain is of fundamental relevance, because it will be a decisive factor in the definition of its priority ("urgency" degree). Usually, the smaller the domain involving a specific problem is, the smaller the individual's interest in solving it will be.

The box "Problems Fields" includes the concepts and individual representations of the experienced problems (the scripts we use to live). It is a kind of "mental index" that classifies lived situations and related them to general concepts "learned" through instruction or experience. As example, the professional of Management that needs to increase the sales through the communication of a sales promotion can classify this problem as belonging to the field "advertisement".

The "Cognitive Universe" is associated to the long-term memory. In it are stored all the lived experiences, as well as the acquired knowledge along the years.

The box of "Emotions" represents the emotional factor, and it is present in all moments of the process: it participates in the prioritization of the actions, in the control of the activities, in the decisions about continuity of the process. It is also influenced by the final solution.

The diagram of the process facilitates the visualization of three types of problems: trivial, whose answer is easily found already in the initial processing, corresponding to the routine situations that we found in the daily life; the difficult problems, whose solution, although no-apparent, can be deduced through the use of a subsequent processing (that Piaget identifies as "reflecting abstraction") that would structure the knowledge without alter the space of researches; and the complex problems.

The problems considered by this article as complex require an enhancement of the research space, because they result of successive inadequate answers as much of the deductive logic as of the previous experiences in relation to the initial problem. That is a delicate moment in the structuring of reasoning, because it is the point that emotions exercise larger pressure, forcing the individual to continually take decisions on the viability of finding a solution or decide to abandon the problem.

In that new, enhanced problem space, enlarged by the impasse caused by the lack of answers of the deductive logic and of the previous experiences, the brain possesses new factors, experiences and abilities, "borrowed" from other areas not considered initially, to apply to the problem, provoking the restructuring of the old mental connections and forming new ones, propitiating the emergence of a majoring balancing: once solved satisfactorily, the problem generates relief and pleasure (in the emotional extent), it produces new experiences (which would be stocked in the long term memory, increasing the individual's cognitive universe), it restructures the fields of problems (through the incorporation of new concepts and representations of problems and possible solutions) and it enlarges the general domain (for the formation of new relationships among fields initially not related).

That modeling facilitates the implementation of artificial intelligence systems that make use or that intend to simulate human creativity. It’s linear approach allows the visualization of the creative process as a flow, especially facilitating its adaptation for use with expert systems, although other techniques can also be used depending on the final objectives of the system.