Machine Learning based Learning Disability Detection using LMS

This paper highlights an E-learning system created using Moodle which is an open-source Learning Management System (LMS) that enables a better learning environment between the tutors and students. This system detects two learner profiles i.e. students with Learning Disability (LD) and without Learning Disability (Non-LD) using dedicated courses designed on the basis of various aspects of an LD student. This work also multiple stages of our approach for informal testing used to capture the learning parameters for Dyslexic students. The first stage i.e. data collection has two approaches where the first approach pertains to a smaller age group of 8-10 years with limited parameters whereas the second approach pertains to the age group 11-13 years i.e. grades 6-8 with more parameters. Natural Language Processing (NLP) has been used to perform Speech-to-Text (STT) conversion on the audio responses of the users. The analysis of these responses have been performed in python language. To detect whether the user has LD (Dyslexia in this case) or not, Machine Learning (ML) is used. Two ML algorithms namely Logistic Regression (LR) and Support Vector Machine (SVM) are used to perform binary classification with LD (1) and Non-LD (0) as the two classes of the dataset. The results are shown for both the approaches and comparative analysis shows that the dataset generated in the final approach for capturing parameters involving NLP is better and more robust. LR algorithm for ML shows better results as compared to SVM for performing detection based on the generated dataset.