Dynamic Sentiment Analysis Using Multiple Machine Learning Algorithms: A Comparative Knowledge Methodology

Human can easily understand or interpret the meaning of language. However, a machine has no natural language to deduce the hidden emotions. Without knowing the context of the word, it cannot simply infer whether a piece of text conveys joy, anger or frustration. Here, sentiment analysis came into picture. Sentiment analysis is the analysis of feelings, attitude and opinions of human emotions extracted from text. It uses natural language processing (NLP) for classifying the text into positive, negative or neutral category. Many businesses nowadays take feedback of the product from the customers to improve the quality or service of the product. Earlier feedbacks were taken by the call center executives but today a vast amount of data is available on the Internet. People share their views regarding products, services, people, etc. Sentiment analysis makes the task easier by extracting the relevant words from the sentences and classifying it in different categories. In this paper, we have described the essential steps used in the process of the sentiment analysis and few fields that work under its umbrella. A comparative analysis of machine learning algorithm like Naive Bayes, SVM, maximum entropy is done along with the few algorithms like artificial neural network and K-nearest neighbor, which can be used in sentiment analysis.