ABSTRACT
Assessment and evaluation are pivotal to any education system. The use of computer-based systems in assessment has led to the automation for objective types of assessment which usually has one correct answer like multiple choice, true/false answers, multiple-response, and matching questions. The educational system has accepted this automation to a certain level at the cost of converting subjective assessment to objective types of assessment. This deployment, in particular, omits the opportunity for assessing the learner’s ability towards application for the cognitive skills of higher levels which are required for constructing a response rather than just remembering the right/wrong options. The current automation in the assessment and evaluation proves to be limited in terms of judging the learners’ cognitive abilities and it can hardly be employed for assessment of learners’ skills and behaviour. The use of Artificial Intelligence (AI), machine learning (ML) and natural language processing (NLP) provides an opportunity to expand the horizon of automated intelligent assessment (AIA) and evaluation from objective to subjective assessments encompassing cognitive, affective and psychomotor domains of learning. The techniques like reading proficiency assessment, automated scoring programs, NLP and auto-grading for short answers are being employed for the cognitive assessment of learners. It is possible to measure the participation and responsiveness in online modules which can be used to assess the attitude and behaviour of learners towards the learning process. The use of computer vision, augmented reality, image processing and AI algorithms can help us to judge the level of performance in acquired skills by the learner. AI based assessments are capable of providing continuous personalized feedback to major stakeholders such as teachers, students and parents about the progress of learners, which can be used to provide help and support as and when needed. This chapter attempts to showcase how various AI based solutions can be employed to automate various types of assessments (predictive, formative and summative) applied to all domains of learning.
