Development Of A Text Classification Model For Tsaravokatra's Customer Complaints Using Naive Bayes And Node.JS

In the ever-evolving landscape of technological innovation that shapes modern business practices, the quest for efficient customer service enhancement becomes increasingly paramount. Within this context, where businesses are navigating the complex terrain of digital transformation to stay competitive, this research sets out to accomplish the specific objective of enhancing customer service to align with contemporary expectations, with a focus on pragmatic and achievable improvements. The primary goal for the study is to navigate the path towards revolutionizing feedback management by developing a Text Classification Model for Tsaravokatra's Customer Complaints Using Naive Bayes and Node.js. Within this endeavor, a monolithic design approach simplifies the intricate processes of testing, code management, and platform deployment. The system adeptly collects and validates user data while categorizing complaints efficiently into distinct categories: Service, Product Quality, and Shipping. The research resulted in the successful development and implementation of a robust text classification model, achieving an 84% accuracy and an impressive F1 score of 84.19%. These metrics signify the model's potential to streamline complaint handling, prioritize issues, enhance response times, and maintain precision and recall balance, with precision at 87.61% and recall at 82.81%. Beyond its immediate application, it represents a notable advancement in text classification, opening doors to innovative possibilities within the field. This research contributes to the ongoing discussion on the transformative power of technology in modern business practices.