Wednesday, January 9, 2013

Text Analytics for the Water Industry

Knowledge management in the Water Industry is archaic, that is if it at all exists. Since the turn of the century there has been dire doomsday predictions of mass retirement of water professionals with 20 to 30+ years of systems knowledge. With the economic slowdown and its effect on public retirement funds, most of those professionals held on to their jobs and now we are seeing the exodus happening. Many water managers are concerned. As their staff retires, they leave with the institutional knowledge and that is quite scary. After all there is no structured method of information capture.
Image courtesy of jscreationzs/
Most of our industry's Operational Intelligence lies unstructured in text-rich data. This data resides in emails, text messages, alerts, notes, chats, reports, and documents (word, excel, powerpoint, database,) etc.  Currently all this data, and hence information, sits virtually unanalyzed and unused. Much of this information  has the potential to deliver operational insights that can be used to make informed smarter decisions in the future.

Text analytics has made great progress in the last decade. Deep language processing capabilities such as summarization, multi-faceted search, and sentiment analysis is mainstream now. Technology is readily available to gather, store, filter, and mine textual information for hidden signals and patterns, trends, and anomalies. And most of this is open-source technology, just the kind of 'fit' public agencies look for.

At PROTEUS, we are working on Text Analytics. With new tools available to handle big data, we can now get access to this unprecedented amount of data and extract value. We no longer have to rely only on models (hydraulic or process) and SCADA information to understand and draw conclusions about our system operations, we can now tap into  a mountain of unstructured and unused complex data and derive insight that can boost our system performance like never before.

PROTEUS model of Operational Intelligence

We believe that once our water clients can crack the code on this cross-enterprise textual data, they will see a significant improvement in operations. Asking the right questions will make all the difference. The more specific and focused the questions can be, the better results can be achieved.

Water managers do not need to do another 'master plan' and set about searching for the data to feed the Analytics engine.  They already have the components to start generating these insights. They can start with existing documents, even with just the organization's email and folders on the server. While every organization is different, we can assure you that results can be seen within just a few months of getting started.

To learn more about how to get the water industry's text analytics initiative underway, on to Smart Operations Intelligence, please contact us at