Sentiment Analysis and Opinion Mining (Business Intelligence 1)

Abstract

This chapter covers the first part of our business intelligence discussion and gives the reader insight into opinion mining and sentiment analysis. Social media are seen as big data in the sense that they provide a massive amount of online reviews and ratings that can be collected and analyzed in order to consider the impact these data may have on organizations. Particularly, several studies have shown that more positive reviews and higher rates for an organization (and its products or services) may lead to a significantly higher number of desired business actions (e.g., higher sales or more subscriptions to an online newsletter). This chapter explains characteristics such as subjectivity and tone in opinions and shows how a sentiment model can be built. The chapter concludes with challenges faced by this research field today. This chapter is primarily situated in the IT department of an organization. This means that the technical execution or implementation of business intelligence techniques will be conducted by IT people or engineers rather than business people. Nonetheless input or involvement of business users is still relevant for successful business intelligence applications because fully automated analyses may lead to inappropriate conclusions or business decisions.