Sentiment Analysis Of Brand Reviews Using Text Blob And Streamlit

Abstract

Sentiment analysis plays a crucial role in identifying and interpreting emotions within textual data such as customer feedback, social media posts, and reviews. This study presents a sentiment classification system categorizing text into neutral, negative, and positive sentiments, aiding organizations in understanding public opinion and enhancing decision-making. To ensure accuracy, the system preprocesses data using cleaning algorithms to remove noise and irrelevant elements.The proposed model employs the TextBlob library for sentiment classification, leveraging its built-in predictive capabilities, while the clean-text library optimizes preprocessing by eliminating punctuation, stopwords, and unnecessary spaces, and standardizing text to lowercase. Key metrics such as polarity and subjectivity assess model performance to ensure reliable outcomes.A Streamlit-based interface enables user-friendly interaction, allowing organizations to extract actionable insights from large datasets. This sentiment analysis tool facilitates improved customer satisfaction, product refinement, and data-driven decision-making.