Stigmergic agent-based adaptive content sequencing in an e-learning environment

摘要

Stigmergic agents (SA) gather their strength from the ant colony metaphor and multi-agent systems (MAS). Ant-based systems provide a means of picking up information through indirect communication by exhibiting stigmergic ant behaviour and MAS lends the benefits of direct interaction among the agents. This paper presents a stigmergic agent framework for adaptive content sequencing (ACS) in e-learning using SA (ACSeLSA). Each learner enrolled in the designed e-course has a personalised stigmergic agent (PSA) that employs an ant-based adaptive content sequencing algorithm, ACSeLAnt, to generate customised content sequence for him/her. By implementing ACSeLAnt, each PSA: 1) exhibits stigmergic behaviour of producing its own pheromone trails and sensing those of other learners to generate the content sequence; 2) elicits recommendations of other similar learners through direct communication with their PSAs and incorporates them into the content sequencing process. The merger is beneficial for knowledge sharing in e-learning systems where there is a dearth of active dialogue among learners.