Optimizing Feedback Recommendation in Smart Training Framework using NLP

This paper primary goal is to assess an item’s resemblance to a ranked group of things. Our research investigates how smart training framework can employ such recommendation systems to provide feedback remarks for students’ open-ended responses. In this case, it’s employed to create a feedback framework that aids educators in evaluating their pupils’ free-form responses. We explore various methods for assessing the semantic relatedness of two answers and use our unique metric, Faculty Assessment points (FAP), to assess the quality of semantic relatedness. Evaluating open-ended data, like natural language, is sometimes regarded as a challenging endeavour. We have developed a software infrastructure that allows us to conduct Randomized Control Trials on the ASSISTments Platform in order to assess the system’s quality.