Learning UML database design and modeling with AutoER

Abstract

Designing models for software systems is a complex activity that takes time and practice to master. Supporting students learning design is time-consuming, and instructors are limited on the amount of feedback they can provide. This work demonstrates a learning system that helps students learn UML modeling to produce database designs. Important features of the system are its ability to auto-evaluate design diagrams based on instructor criteria, and automatically generate question variants allowing students to practice on their own. The system is integrated into the PrairieLearn learning management system allowing for its deployment at other institutions. The demonstration shows the random question generation, how students construct database designs, and the auto-marking capabilities. The system and practice questions are publicly available for instructors to use to help students learn design modeling.