Sentiment Analysis of Makeup Alley Website Cosmetic Reviews Using SVM and TF-IDF

This paper focuses on performing sentiment analysis of cosmetic reviews using a machine learning algorithm. The study aims to classify reviews into positive, neutral, and negative opinions using aspect-based sentiment analysis with TF-IDF and SVM algorithms. The research questions include understanding user opinions on cosmetic products and identifying the most common positive and negative sentiments in reviews. Data cleaning, tokenization, stopwords removal, normalization, and lemmatization techniques are employed to achieve these goals. The study uses the Makeup Alley website as a data source and employs web scraping to obtain the data. The Natural Language Toolkit (NLTK) is used to simplify the removal of stopwords. The research seeks to document the step-by-step process of developing a sentiment analysis application tailored to cosmetic product reviews.