The coming democratisation of emotions analytics

Dan Somers
DOI: https://doi.org/10.69554/edri2633
2019-02-06
Abstract:The ability to understand emotion from human communication is essential for marketers and data analysts, particularly in a world of growing non-verbal communication, such as reviews, complaints, surveys social media and chatbots. Thanks to advancements in artificial intelligence (AI) text analytics, a lot of information typically missed in sentiment analysis can now be analysed and early warning of issues flagged much earlier. This paper looks at how AI text analytics will succeed sentiment analysis and how it can identify additional customer emotion and intent by using concepts rather than simple keywords in its analysis. The paper also looks at how the most successful AI text analytics feature human-in-the-loop technology that optimises the combination of automation and human involvement. This makes the need for data scientists to run day-to-day models obsolete as the human input can come from the business user instead. The paper presents examples of how AI text analytics and emotions analytics can be applied, and argues that AI text analytics can reduce reliance on data scientists, provide early warnings of issues and analyse a much deeper level of sentiment and customer intent with higher levels of accuracy.
What problem does this paper attempt to address?