Optimizing Quality of a System Based on Intelligent Agents for E-learning

Distributed Artificial Intelligence is a subfield of artificial intelligence systems, that aims to build intelligent agents, that can make decisions in order to achieve their goals and are set in a world populated by other intelligent agents (artificial or human) who in turn have their own purposes. In this paper we have presented intelligent agents for e-learning. From quality perspective, the system presented indicates a way to search for concepts in a distributed e-learning environment. Based on the requirements of ISO 27000, we reviewed the stages for creating the system. Initially, higher ranking agents interrogate the server's local knowledge base agents without modelling them. Then, the agents shape the knowledge bases from the servers to avoid asking the same question twice if the local agent does not know the answer. Further, it introduced the possibility for agents to cooperate, so that they can ask questions of each other if they are on the same server. Finally, priority queues were implemented on the servers, so that only a fixed maximum number of agents could be served at a time, in a descending order of priorities.