Vol. 7, Issue 1, Part I (2025)

Sentiment analysis has emerged as one of the fastest-growing areas in computer science, driven by the increasing importance of understanding consumer opinions in the digital era. As advancements in Natural Language Processing (NLP) and machine learning continue to evolve, sentiment analysis plays a pivotal role in deciphering emotions, opinions, and attitudes expressed through textual data. This research leverages a dataset of product reviews collected from Flipkart.com, one of the largest e-commerce platforms. By analyzing the ratings provided by customers, the study performs a comprehensive sentiment analysis to categorize reviews as positive, negative, or neutral, offering insights into the general perception of various products.

The methodology includes preprocessing the dataset to clean and normalize the text, followed by sentiment classification using state-of-the-art machine learning models. Additionally, the research identifies the most frequently used words and word pairs (bigrams) to reveal significant patterns in customer feedback. A word cloud visualization is used to highlight the most prominent terms, while count plots are employed to examine the distribution of sentiments across different product categories. The study also explores the relationship between sentiment polarity and product ratings to understand how specific words correlate with customer satisfaction.

The findings suggest that identifying the most critical negative and positive words within the dataset can offer valuable insights into the psychological state of consumers. This understanding can not only help businesses enhance their customer service and marketing strategies but also provide a deeper look at consumer behavior and decision-making processes. Moreover, the study emphasizes the potential of sentiment analysis in the e-commerce domain, where understanding customer sentiment can significantly influence product development, sales strategies, and brand loyalty.