Meta-level Representation for Enhancing Adaptibility for Multi-tier Users in Web-based Modeling System

초록

A meta-level representation of an adaptive modeling system (AMS) is proposed by extracting relationships between models on web-based system. Adaptive behaviors of the AMS are identified based on existing study into adaptiveness of modeling system for multi-tier users. A variety of knowledge, which enables the system to adaptively behave, is well known. The concept of three layers, which are Dialogue Generation, Chunk Extraction, and Relationship Investigation, of modeling system is adopted to achieve the goal of the adaptive modeling. The Dialogue Generation layer proceeds the modeling of users’ problem. The Chunk Extraction layer handles meta-level knowledge, i.e. subordinate relationship between models. Finally, the Relationship Investigation layer manages diverse models in the system and generates the subordinate relationships among the models.