Abstract:
The business environment of today has grown incredibly competitive and difficult. Growth
of businesses now places a lot of emphasis on customer happiness. To understand and
meet the demands of their clients, business organizations devote a significant amount of
money and human resources to different techniques. But many businesses are failing to
satisfy customers as a result of the manual analysis of customers varied wants being done
in an imperfect way. As a result, they are losing their customers' trust and increasing their
marketing expenses. Sentiment Analysis is a solution that we can use to resolve the issues.
Machine learning (ML) and natural language processing are both included into the system
(NLP). Analysis of people's feelings about certain topics, products, and services is called
sentiment analysis, and it is used rather often to get insights into how the general public
thinks about certain topics, products, and services. We are able to do that by using any
data that is found online. In this article, we present two natural language processing
strategies (Bag-of-Words and TF-IDF) as well as various machine learning classification
techniques to perform sentiment analysis on a large dataset that is imbalanced and contains
many classes of data.
