Sentiment Forecasting in Women’s Fashion E-Commerce: A Machine Learning Perspective

Sentiment analysis is a technique that uses natural language processing to determine whether a piece of text has a negative, positive, or neutral sentiment. It is also known as emotional AI or opinion mining. It proves invaluable when dealing with noisy, unstructured customer data from diverse channels, where manual processing is impractical. This technique is widely applied across social media posts, blogs, surveys, reviews, and news articles to derive sentiment scores and extract valuable insights. The study in question focuses on analysis of sentiment in the field of reviews for women's clothes, applying Naive Bayes, support vector machine, neural networks, logistic regression, and other machine learning approaches. Evaluation parameters similar to precision, accuracy, F1-score, recall, and (AUC) area under the curve are employed to assess the model performance. The research concludes that logistic regression and neural network models yield the best results, achieving a remarkable 92% accuracy. Notably, Naive Bayes demonstrates efficiency, particularly in handling larger datasets. This study contributes valuable insights into sentiment analysis, classifying women's clothing opinions as positive, neutral, or negative with practical implications for predictive modeling.