Surveying and Evaluating Artificial Intelligence in Automated Assessment Systems for Pen-and-Paper Tests

Abstract

A huge portion of the testing process in most schools is still conducted in a traditional pen-and-paper test manner. This is not only the case in huge courses, large organizations and massive testing events that involve thousands of candidates, but also in smaller schooling communities with insufficient personal computers and human resources. These paper tests are mostly examined manually by the teaching staff, which imposes a significant burden on them. Hence, there was a need for any kind the grading process automatization that would accelerate the assessment process and disburden the teaching personnel. Therefore, software systems for automated assessment of paper tests were developed, partially or fully aiding the teachers in the examination process. These software systems already exhibit some form of artificial intelligence behavior. Artificial intelligence is already providing enormous opportunities for this type of software, and it is simply a matter of time before this software will grade all kinds of tests on its own without human supervision. Due to its recognized importance, this paper provides a detailed analysis and review of available software systems that can be used when assessing pen-and-paper tests.