Sentiment-Based Insights Into Amazon Musical Instrument Purchases

The project aims to analyze Amazon music instrument reviews to understand customer satisfaction in terms of key sentiments related to the products. Online shopping has made customer reviews increasingly relevant for both buyers and sellers. The project analyzes unprocessed text data using Natural Language Processing (NLP), which includes steps like text cleaning, tokenization, stop-word removal, stemming, and lemmatization. Word embeddings or term frequency-inverse document frequency (tf-idf) are used to encode the large dataset. A method of sentiment analysis on user reviews is described, with the dataset going through preliminary NLP work. Sentiments are classified using various models, evaluated based on accuracy, cross-validation scores, and classification reports. Visualization graphics depict sentiment distribution, emphasizing the potential of NLP for better sentiment analysis performance.