Automated assessment of free text questions for MOOC using regular expressions

Abstract

Presently, most platforms used on the selection of Massive Open Online Courses (MOOC) available online have various automated methods of assessment. These type of tools are based on applications that analyze the answers using a pre-correction algorithm. Most of these programs run several types of automatic assessment, but the possible use of the technology for each of them differs with respect to the kind of automation applied. The role of technology in the objective test for online education has become extremely common, so it can be found in various MOOC platforms with this type of questions or quizzes, because the assessment system can be fully computerized (from the test design to its correction and reporting). However, not all of the assessment instruments can be easily implemented in automatic mode with the use of technology. This paper seeks to research and clarify a type of assessment tool in which the use of technologies is quite low, namely, the essay question type and within them, the short answer question type or free text question type, using regular expressions. The large number of students who would be in MOOC prevents a teacher from assessing responses of thousands of students in a finite time without the aid of technology. This research analyzes the results of an MOOC from hundreds of students to verify that the use of regular expressions in an MOOC platform is not only recommended but also necessary.