Abstract
The rapid growth of educational applications have reshaped digital learning that provides a wide range of tools to users globally. There are multiple platforms for these learning applications i.e., google play store, web store, apple app store etc. This thesis presents a systematic approach to analyze user feedback from educational applications such as Udemy, Khan Academy, Duolingo, Solo Learn etc., available on google play store. It utilizes Natural Language Processing (NLP) and Machine Learning (ML) techniques for comments filtration based on user centric requirements. The main purpose of this research is to derive useful information from user sentiments, preferences, and challenges by analyzing user comments. The objective is to enhance app quality and overall user experience. User reviews serve as the core data source for analysis. In our thesis this data provides valuable input for software requirement gathering which is a crucial phase in the Software Development Life Cycle (SDLC). Different ML Algorithms are used in this research to effectively filters and categorize user comments. These filtered comments can be used as significant user-centric requirements. These insights enable developers to optimize existing app features, remove redundant functionalities, and tailor applications to better meet user requirements. The study further evaluates the proposed methodology by comparing it against traditional requirement gathering methods. The F1 Score, precision, recall are employed as a performance metric to assess the accuracy and efficiency of the filtered comments. Random Forest and XG Boost showed 96 %, SVM showed 95 % , Logistic Regression showed 93% and Naïve Bays 91% accuracy. XG Boost hence outperformed other classification algorithms approaches by attaining maximum accuracy, precision , recall and F1 score. XG Boost is hence proposed approach for doing this classification task of user centric requirements gathering from user comments.
Replies