System for automatic generation of algorithm visualizations based on pseudocode interpretation

Abstract

Algorithm visualization systems have not been as widely adopted by computer science educators as it might be expected from the firm belief that they can enhance computer science education. Two key impediments for widely adopting AV technology in mainstream computer science are: effectiveness and enhancements of learning with visualization and effort needed to create algorithm visualizations. In this paper, we present the interpretation based system capable of automatic creation of algorithm visualizations by interpreting unmodified algorithms written in pseudocode. Although system is interpreting unmodified source code (code without any annotations for triggering appropriate visualization routines), due to the ability to automatically detect interesting events system is able to create visualizations at a sufficiently high level of abstraction so that the emphasis is on algorithmic conceptually relevant principles. Providing users with full control over input data set and by accompanying animation with explanatory messages, highlighting currently executing pseudocode line and providing possibility to inspect variable values at any step visualizations created by our system that can enhance learning and help students mastering algorithms basic concepts.