Regularized neuro-fuzzy AI model to aid score management in Online distance learning forums

This paper proposes to use an artificial intelligence (AI) model based on neuro-fuzzy techniques to aid in the automatic evaluation of notes originated from the student's interaction with their performed activities in online distance learning (ODL) forums. The evolution of non-classroom teaching allows new business opportunities and studies to emerge for the population. Some people who do not have enough time to attend traditional teaching choose distance learning to boost their tasks and money. The increasing demand for ODL courses creates challenges that are mainly aimed at automating the tasks commonly required for students' evaluation routines during their activities. For distance learning to maintain an acceptable level of costs, routine activities must be automated to reduce values related to more straightforward operations execution. In this paper, we will use a real dataset on evaluations of frequent interactions of academic in forums, allowing the obtained data to be submitted to a fuzzy neural network able to estimate the value of the student's score value according to the activities carried out by them, beyond the extraction of knowledge through fuzzy rules. Model outputs confirm that the approach may be feasible to automate the presence and participation process in ODL forums through a specialist system based on fuzzy rules. The tests performed with a resulting low RMSE of 1.37 suggest that our neuro-fuzzy-based AI approach performs better than traditional state-of-the-art regressive models.