Towards the Automatic Recognition of Computational Thinking for Adaptive Visual Language Learning

Visual programming languages can be used to make computer science more accessible to a broad range of students. The evaluative focus of current research in the area of visual languages for educational purposes primarily aims to better understand motivational benefits as compared to traditional programming languages. Often these visual languages claim to teach students computational thinking concepts; however, although the evaluations show that students may exhibit more enthusiasm, it is not always clear what computational thinking concepts students have actually learned. In this paper we attempt to develop a visual semantic evaluation tool for student-created games and simulations that goes towards depicting the computational thinking concepts implemented by the students. Through semantically analyzing a given student's created projects over time, this visual evaluation tool, called the Computational Thinking Pattern (CTP) graph, can possibly indicate the existence of computational thinking transfer from games to science simulations.