B

Architecture à tableau noir

BBA

Un modèle informatique où plusieurs agents partagent une base de connaissances commune pour résoudre des problèmes de manière collaborative.

Tableau noir est un modèle informatique largement utilisé dans is a computing model used in intelligence artificielle where multiple agents or processes collaborate to solve complex problems by sharing a common base de connaissances, referred to as the ‘blackboard.’

In this architecture, the blackboard acts as a shared workspace where different knowledge sources can contribute information, hypotheses, or solutions. Each agent or module, known as a ‘knowledge source,’ has its own expertise and can read from and write to the blackboard. This allows for a flexible and dynamic problem-solving approach, as different agents can contribute at different stages of the process.

The blackboard architecture is particularly effective in scenarios where problems are too complex for a single agent to solve independently. For example, in traitement du langage naturel, various agents might handle syntax, semantics, and discourse analysis, contributing their findings to the blackboard to create a coherent understanding of the text.

Composants typiques d'une architecture à tableau noir :

  • Sources de connaissances : Le référentiel central où sont stockés les données, hypothèses et solutions.
  • Composant de contrôle : Independent agents or modules that access the blackboard, contribute knowledge, and make inferences.
  • Composant de contrôle : Manages the activity of knowledge sources, determining which source can act and when, often based on the current state du tableau noir.

Cette architecture est polyvalente et a été appliquée dans divers domaines, notamment reconnaissance vocale, image processing, and expert systems, showcasing its ability to handle complex, multifaceted problems through collaboration.

oEmbed (JSON) + /