Guide to big data analytics tools, trends and best practices
A comprehensive collection of articles, videos and more, hand-picked by our editors
While big data analytics is still new enough to evoke the aura of the Wild West, some companies have been at it long enough to develop true competencies. To capture and benchmark the evolving best practices of such companies, The Data Warehousing Institute is undertaking a new study, according to Fern Halper, author and research director for advanced analytics at The Data Warehousing Institute.
Early patterns are emerging. Halper said people addressing big data analytics tend to come at the problem from two distinct directions. Traditional companies focus one way, while Internet-based firms and smaller companies focus in another.
"The Internet companies are all about getting something done; they are about moving as fast as they can. That can be good, but it can come back to bite you," she said. "The larger companies are moving slower; they may build a proof of concept and it may take longer to permeate the organization."
Data managers at Internet or Web applications companies know they ultimately need to put more formal processes in place to handle big data analytics -- corporate compliance is often a driver -- but the shift in direction can be hard to manage if they are growing rapidly, according to Halper.
Internet companies are about … moving as fast as they can. That can be good, but it can come back to bite you.
research director, TDWI
Data managers at traditional enterprises also face hurdles moving big data analytics from initial deployments to wider uses. "It is hard to get from narrow, early adoption to broader corporate adoption for big data analytics," she said.
Maturity model has five dimensions
Halper, who earlier this year co-authored Big Data for Dummies with Judith Hurwitz, Alan Nugent and Marcia Kaufman, said The Data Warehousing Institute (TDWI) looks at big data analytics as the application of analytical techniques not just to big sets of data, but also to highly varied types of data. In fact, data shops that are adept at handling large data sets in transactional and relational form still have a lot to learn when they take on more varied information types, such as unstructured data, as part of new initiatives.
As part of the effort to create TDWI's Big Data Maturity Model, Halper talked with users and vendors, looking at five specific dimensions of the big data maturation process: organization, infrastructure, data management, analytics and governance.
She said a benchmark survey will be published on the TDWI site in October, and users will be able to go in, answer questions, and see where they stand relative to their peers in terms of maturity of big data analytics.
Jack Vaughan is SearchDataManagement.com's news and site editor. Email him at firstname.lastname@example.org.
Follow SearchDataManagement.com on Twitter: @sDataManagement.