NLP uses in BI and analytics speak softly but carry a big stick

When self-service BI went mainstream, it exploded the paradigm that specific technical skills were required to build reports from data sets. Now, ease of use and simplicity are the expectations for BI tools.

The latest extension of that trend is natural language processing (NLP)-based querying. BI and self-service analytics vendors are increasingly adding NLP uses to their platforms to make querying a database more like a web search. The idea is to let any line-of-business user explore a data set and find answers for themselves using simple, intuitive search terms.

Gartner has called augmented analytics, which uses machine learning and NLP to support simpler data exploration, "the next wave of disruption in analytics and BI," and said the AI technology will be the most important factor driving new analytics and BI purchases by 2020.

But that doesn't mean you can talk to your data set like you'd talk to another person. There are still fairly substantial limitations on what kind of questions you can ask BI tools that have NLP uses built in, and understanding what features are contained in a data set is still important.

That's why, as we explain in this handbook, it's critical for departments implementing self-service analytics tools with NLP features to set expectations with users. We also examine some of the more impactful NLP use cases and look at how conversational NLP is getting closer to tapping into user sentiment and emotion.

By

• Ed Burns