Sentiment Analysis using Machine Learning

Abstract

With the rise of e-commerce, consumers are becoming accustomed to shopping online and leaving reviews of their experiences on retailer and review websites. These op-ed pieces provide future customers with useful information for making decisions, and they also help businesses improve their goods and/or services. Yet, when the number of reviews increases quickly, users are forced to deal with a serious information overload issue. Several opinion mining techniques, such as opinion summarization, opinion polling, and comparative analysis, have been suggested to address this issue. How to correctly estimate the sentiment direction of review sentences is the main challenge. Sentiment analysis is a quickly developing subfield of the study of Natural Language Processing (NLP).In recent years, it has drawn a lot of interest. Sentiment analysis is used to assess or investigate user comments in order to make judgements about their opinions. Sentiment analysis is a machine learning (ML) technique in which computers classify and investigate human attitudes, feelings, and viewpoints towards the things that are communicated through text, star ratings, thumbs up and down, and other textual or graphical representations. The data for this study came from online product reviews acquired from the model website we created. Adjectives and adverbs can convey the opposite feeling with the use of negative prefixes. Such words are located using a negative phrase identification method. By using evaluation measures, the performance is assessed. Finally, we provide insight. Finally, we also provide a preview of our upcoming sentiment analysis research.