A Study on Text Classification for Applications in Special Education

Text classification, or text categorization, is the process that categorizes text into organized groups. Text classifiers can parse text and categorize it based on its content. The text analysis is done using Natural Language Processing (NLP). In recent years we have much more understanding to learning disabilities and how to diagnose and categorize their impact. In this paper we studied learning difficulties as a classification problem and use the NLP tool to solve it. In addition, the results of classifiers such as Naive Bayes, SVM and CNN are compared. The precision results show that the Naïve Bayes classifier had the lowest accuracy of 71.72%, compared to the other two, where the CNN classifier has a 82.79% accuracy and the SVM has the best accuracy in percentage of 93.03%.