What NASA and OdinText Have in Common
Word Clouds Don’t Fly at NASA Either
This year I participated on the closing panel at the Text Analytics Summit in Boston. The most interesting presentation at the event was one given by NASA’s Ashok Srivastava on how they are using text analytics to make air travel safer in the US, and in fact around the world, by monitoring what pilots and ground staff report in regard to various safety and mechanical issues.
Using text analytics in this way is obviously a bit different than how we at Anderson Analytics use our software OdinText. Understanding how different consumers view products and the problems these groups experience requires a different approach with different risk tolerances. While I believe understanding VOC to create the best products and services is very important, imagine the complexity and inherent danger involved in the system NASA analyzes.
Text Analytics Objective: Make Air Traffic Safer
[24 Hour System View]
After the presentation I spoke to Ashok and found it interesting that while we obviously have different objectives, data and therefore approach text analytics somewhat differently, one of the several areas that our approaches have in common is data visualization. It has always surprised me how many ‘text analytics’ firms use simplistic word clouds as output; one of my many pet peeves in our industry. I wasn’t surprised that NASA didn’t use them either. Instead I found that we both include correlations of unstructured data and handle these with very similar statistical/visualization techniques for insights far more meaningful than word clouds.
Ahead of Text Analytics Summit West in San Francisco next month, Text Analytics News contacted us both for an interview on how we think text analytics and big data can become a default business solution. Obviously the ROI needs to be better understood by many companies before even greater adoption takes place. In the private sector, this means there is still a first mover advantage, and firms who implement text analytics early can gain an information advantage over their competition.
Again, industry and domain specifics aside I was pleased to see we are in agreement on many of these issues as well. You can read the Q&A on the Text Analytics News website here.
PS. I understand Ashok will be speaking Text Analytics Summit West as well. You can get a glance at his presentation here: