Evaluating AI Algorithms for Sentiment Analysis in Customer Feedback Systems
Dr. P U Anitha, Borra Srinath, Patchineela Sudheer Kumar
This research evaluates various AI algorithms for sentiment analysis in customer feedback systems,
focusing on both traditional machine learning and advanced deep learning models. We analyzed the
performance of Support Vector Machines (SVM), Naive Bayes, Decision Trees, Convolutional Neural
Networks (CNNs), and Long Short-Term Memory (LSTM) networks using a dataset of 100,000 customer
feedback entries. The evaluation metrics included accuracy, precision, recall, F1 score, processing
time, and scalability. Our results indicate that deep learning models, particularly CNNs and LSTMs,
outperformed traditional machine learning models in terms of accuracy and recall, with LSTMs
achieving the highest overall performance. However, these models also required significantly more
processing time and showed varying scalability. In contrast, traditional models like Naive Bayes and
Decision Trees demonstrated faster processing times and higher scalability but with lower accuracy.
This study provides a comprehensive comparison of these algorithms, offering valuable insights into
their effectiveness and practical applicability in real-time sentiment analysis.
