A Comparative Study of Customer Sentiment Analysis on Amazon Wi-Fi Routers Using Machine Learning Models

Customer sentiment analysis is essential in understanding consumer perceptions and preferences, thereby enabling businesses to make prudent decisions. The paper presents a comparative study of customer sentiment analysis on Amazon Wi-Fi routers using various machine learning models. With the increasing importance of online reviews and ratings, accurately assessing customer sentiments becomes crucial for manufacturers and retailers alike. A large dataset of customer reviews on Amazon for different Wi-Fi router models is collected that consisted of a diverse range of sentiments expressed by the customers, including neutral as well as positive and negative opinions. Data preprocessing methods such as tokenization, stemming, etc., were employed to make sure that there is robustness in the data used. Further, assessment of several machine learning models and their comparison for sentiment analysis, combining of traditional models, were done. An overarching complete analysis of these models based on various evaluation metrics, including accuracy, precision, etc., was then performed. An investigation into how different feature extraction techniques and hyperparameter tuning affect the effectiveness of these models was done. The experiments aim to identify the most efficient approach for sentiment analysis on Amazon Wi-Fi router reviews. Also, the results demonstrate that the Light Gradient Boosting Ensemble model outperforms traditional machine learning models, yielding higher accuracy and improved sentiment classification. Our findings suggest that businesses operating in the Wi-Fi router industry can leverage ML models, specifically Light Gradient Boost-based approaches, to gain deeper insights into customer sentiment, which can help companies identify potential areas for product enhancement, customer satisfaction, and develop effective marketing strategies.