Artificial Intelligence for Smart Customer Feedback Management: Improving Content Analysis Approaches

In today’s information-rich environment, organisations face the challenge of efficiently processing and interpreting large volumes of textual data, particularly customer feedback. This paper presents the design and implementation of an AI-powered web application for automated text classification using a neural network approach. The system is developed to support businesses in analysing customer reviews, identifying sentiment, and facilitating prompt responses to issues that may impact customer satisfaction. The classification model is based on an LSTM neural network, trained on a dataset of 2,500 synthetically generated and pre-labelled reviews. Model training was conducted using the Google Colab platform, achieving a classification accuracy of approximately 94%. The backend of the application is built using Python, Flask, and MongoDB, providing robust data processing and storage capabilities. The frontend, developed with Node.js and React, offers intuitive user interaction features, including data import, visualisation of review statistics, and filtering by sentiment. The application enables automatic classification of reviews into positive or negative categories, real-time graphical reporting, and prioritisation of negative feedback for managerial attention. By automating the feedback analysis process, the system enhances operational efficiency, reduces manual workload, and supports data-driven decision-making. The proposed solution demonstrates the practical value of AI and machine learning in improving customer relationship management and service quality in both B2C and B2B contexts.