Tom Davenport on Taming the “Data Plumbing” Problem

 

WSJ CIO 2

 

How Tamr’s Unique Machine/Expert Approach to Curation Is Helping Solve the Data Plumbing Problem

Data Scientists or Data Plumbers? In his WSJ CIO Journal piece, “Taming the ‘Data Plumbing Problem,'” Tom Davenport says he likes “the idea that Tamr uses two separate approaches” — machine learning and expert insight — to solve the plumbing problem through data curation.

“If the data problem is somewhat repetitive and predictable, the machine learning approach will develop an algorithm that will do the necessary curation.  If the problem is a bit more ambiguous, the crowdsourcing approach can ask people who are familiar with the data (typically the owner of the data source) to weigh in on its quality and other attributes.”

Davenport goes on to cite how Thomson Reuters, Novartis and Gloria Jeans are already benefitting from Tamr’s combined solution to cleaning and integrating data through curation.