The Essential of Sentiment Analysis and Opinion Mining in Social Media : Introduction and Survey of the Recent Approaches and Techniques

Abstract

With evolution of social network and Web 2.0, people not only consume content by downloading on web but also contribute and produce new contents. People became more eager to express and share their opinions on web regarding daily activities as well as local or global issues. Due to the proliferation of social media for instance Facebook, Twitter, Youtube and others, sentiment analysis and opinion mining grow rapidly. It branches out from the field of natural language processing and data mining particularly from web mining and text mining. Why sentiment analysis and also known as opinion mining is prevalent and relevant nowadays? When we try to decide to purchase a product, we are likely to get the opinions from friends or relatives and do some surveys before we purchase the product. Hence, opinions are undeniably the key influencer of our behavior as well as the central to nearly all of the activities. Within the opinions, we often find the neutral, positive and negative polarities in the sentences. Based on the sentiment analysis taxonomy, it has opinion mining to have the opinion polarity classification, subjectivity detection, opinion spam detection, opinion summarization and argument expression detection. On the other hand, emotion mining has the emotion polarity classification, emotion detection, emotion cause detection and emotion classification. If it is based on granularity level, it has sentence level, document level and aspect/entity level of sentiment analysis. As for the machine learning approaches, it has semi-supervised learning, unsupervised learning and supervised learning of sentiment analysis.