Toward Intelligent Solution to Identify Learner Attitude from Source Code

Abstract

Technology became widely used in all teaching processes. In some advanced applications, teaching and learning are operated in virtual environments allowing consequently developing unlimited new paradigms. Programming and software engineering practical works are generally executed in Integrated Development Environments. This last contain mainly students’ source code. Analyzing source code is usually processed to verify syntax correctness or to verify solution relevance to problem issues. Student personality dimension is an important factor that teacher undertake during teaching mission. This last can serve in understanding student nature and adapting teaching pedagogy to his personality. This paper aims to explain the idea of gathering insights from student source code. We explain how we can use mainly Machine Learning and Natural language processing to obtain a kind of source code understanding. As a first step, which is the purpose of this paper, we have built an intelligent solution architecture that shows how to collect source code rich data and go through analysis technics to finally use machine learning model to learn about this domain.