Self-learning system with automatic feedback for text answers

Abstract

This paper presents details about a system developed using Natural Language Processing tools and methodologies to automatically evaluate a text answer by comparing the semantic similarity between the model answer with the provided student answer. System generates a score according to the matching percentage of the semantic similarity using the assigned marking pattern for the question. This system is embedded in a web application to be provided as a service for students and teachers to promote self-learning through question answering.