Adaptive Learning Objects Assembly with compound constraints

Abstract

This article addresses how to fulfill ALOA (Adaptive Learning Objects Assembly) which provides users personalized learning resources and learning path based on evolutionary PBIL (Population Based Incremental Learning) algorithm. Both the users' preferences and learning resources' intrinsic characteristics are considered here. And the experience from proficient experts is used to give the LO (Learning Object) difficulty level and important grade which guides the LO's sequencing and selection. The constraints of knowledge such as basic ones, itinerary ones and compulsory ones are also vital factors for ALOA. All of above are modeled as a Constraint Satisfaction Problem (CSP). The PBIL algorithm is proposed and applied to ALOA firstly. The hybrid intelligent evolutionary algorithm is tested on true teaching data and the participants also give the learning feeling. We also obtained the experiment data from the tested data and questionnaire. ALOA's good validity, accuracy, and stability performance are verified.