Abstract:

The implemented solution not only delivers deeper, more actionable insights in a scalable way but also provides the option to reduce survey time. We found that its automated text categorization can predict overall satisfaction even better than quantitative driver questions. The presentation will show that first, it needs a thoughtful approach to NLP text analytics to achieve a coding with acceptable (human equivalent) accuracy and second, it needs flexible machine learning techniques to meaningfully understand the impact of themes. In the presentation, we will also talk about our approach to converge business intelligence with automated text analytics techniques to provide an easy to use and effective solution to drive action.