Main Article Content
There is a wide range of unstructured data available on social networking platforms. There is an urgent need to classify this data. Sentiment analysis helps in classifying this data beyond numerical methods, focusing more on qualitative analysis of data. The aim of sentiment analysis is to acquire the abstract emotions and opinions from data which helps in making appropriate decision regarding the user’s intention behind a message. Sentiment analysis (opinion mining) assists in evaluating the data via means of different indexes and metrics. Social network analytics makes it possible to consider data as links between users which aids opinion mining. The paper deals with three important concepts to understand the interactions between users on social networks namely tie strength, homophily, and source credibility. Sentiment analysis also faces many challenges particularly posts involving sarcasm and irony which are extremely difficult to analyze. Opinion mining involves three levels of analysis including message level, sentence level and entity and aspect level. Opinion mining is a broad field which has a great scope of research in the upcoming future. There are many machine learning techniques which can be used for analyzing sentiments however this paper mainly discusses utilization of social network analytics for sentiment analysis.