Adaptive multiagent system for learning gap identification through semantic communication and classified rules learning

Abstract

Work on intelligent systems application for learning, teaching and assessment (LTA) uses different strategies and parameters to recommend learning and measure learning outcome. In this paper, we show how agents can identify gaps in human learning, then the use of a set of parameters which includes desired concept, passed and failed predicate attributes of students in the construction of an array of classified production rules which in-turn make prediction for multipath learning after pre-assessment in a multiagent system. The context in which this system is developed is structured query language (SQL) domain with concepts being represented in a hierarchical structure where a lower concept is a prerequisite to its higher concept.