Machine Learning based Code Assessment Systems
Bharati Ainapure, Reshma Nitin Pise, Digvijay Singh · 2022
Today, there is a huge demand for automatic assessment of computer programs due to an outsized variety of Massively Open Online Courses-MOOCs or online courses. Assessment of computer programs is additionally beneficial to the businesses during their recruitment procedure, where participants are required to solve a series of programming statements. As a result, computerized grading of open-ended responses is becoming a hot topic in academia. In existing techniques assessment of computer programs is carried out based on the scores computed by passing the code through a variety of test scenarios. This does not reflect the programmers’ overall abilities. Therefore considering the programming style of the programmer, the code evaluation can be made more accurate and easy by combining existing systems with machine learning methods. MOOCs are fundamentally an asynchronous platform and a process for teaching using pre-recorded classes, resource video files, class notes, assignments, and tests, all of which are usually available online and provide self-assessment at frequent intervals throughout the learning process. In this paper machine learning based code assessment system is proposed for programming languages.