Geeks Wanted -- Big Data Firms Push Data Scientist Development

Teradata TDC +0.00%, IBM IBM -0.65%, SAS and Tableau are among the big data specialists offering training for data scientists, sophisticated analysts, business intelligence experts –  whatever they end up being called, but people who can work with the data they can collect and store.

Teradata said it wants to ensure there are enough data experts to make good use of its technology. Companies are working with universities on data analytics programs. They are also working on certification programs to provide credentials for geeks who might not want to go through four years of sociology and readings in French literature on their way to getting a job in big data.

According to Teradata, job prospects in business intelligence or analytics are plentiful. It supplies university programs with large data sets. (See my story on how SAS worked with Weslyan.)

Barb Wixom, associate professor of commerce at the University of Virginia’s McIntire School of Commerce, said that faculty are looking for access to real, big-data sets.

“They want to show students the impact of the data explosion, demonstrate the linkage between data and business outcomes, and teach exactly how to achieve those outcomes,” she added. “Corporations are making available real-world cases and big data sets, and through academic alliance programs like the Teradata

University Network(TUN), they are working with professors to develop meaningful analytics assignments,

teaching notes and other pedagogy so that data sets are consumable by professors and students.”

Alan Chow, chief customer officer at Teradata, said big data requirements come on top of basic analytics and understanding of statistics.

“We are working with universities to add skill sets in Hadoop and MapReduce, but from a management and business standpoint,” he said.

Thilini Ariyachandra, associate professor of management information systems at Xavier University in Cincinnati, Ohio, said the university network is led by academics so it knows what university instructors need.

Teradata has organized three business intelligence conferences in 2009, 2010 and 2012 to bring together academicians, industry experts and data professionals to figure out what is needed in business intelligence and big data teaching. Business intelligence programs are increasing in universities — the number grew from 15 programs in 2010 to 131 in 2012.

“The academic world tends to be a bit slow when it comes to meeting demands,” said Ariyachandra. “Our survey showed there is a huge demand and a big gap in the supply of skills. Academia is trying to catch up.”

Big Data And Business

Business intelligence might fall into MIS or computer science courses in schools, but with big data there is more collaboration going on with finance, marketing, statistics, computer science and management information systems, said Chow.

“These areas are coming together and sometimes co-teaching courses to make sure students get the right skill sets that are needed.” Other companies, including IBM and SAS, are building academic alliances to get the right skill sets from university programs, he added. Employers are also looking for experience in addition to degrees, so internships and semester-long projects will help a new graduate land a job because they will have some of the applied skills needed.

Teradata has a close working relationship with faculties in the San Diego area,” added Chow. “We provide support in terms of resources and we are working with partners to provide software packages for academic use to help the faculty develop course-ware. And the Teradata University Network  has a great network for faculties to share their course-ware and other materials.”

Teradata provides the university network with the company’s view of the business needs and industry trends, he said, and then the instructors and professors decide how they want to respond.

Ariyachandra said employers want the soft skills, such as communication, but they also see a need for better query and SQL skills and basic analytics data management. Many Teradata customers want to join the Teradata University Network board so they can explain what they want from graduates.

“Teradata customers want to reach out to the academics; that is one of the big things we are looking at through the TUN board,” she added, “connecting the customers with students going forward to help with the gaps in knowledge and skills.” Recent surveys show that students realize their courses in business intelligence would be useful.

“We saw a sense of enthusiasm; they said they were taking the courses because they wanted to and found them interesting,” said Ariyachandra. “Usually it is hard to find students to say they are excited about a topic. I think there is a lot of confidence in analytics as a field. They know there is a big gap and they are coming to this field quite willingly to learn more.”

Educating Managers

Teradata also offers a certification program, separate from its academic support programs, said Chow. Now, he added, Teradata is looking to add some education for managers.

“We have noticed that management currently in industry lacks some of the necessary skills to understand how to use big data for their companies. It would help to have some executive programs to train existing managers on how to use big data.”

GE

GE is recruiting about 400 data scientists or transferred them into a data science group from other parts of the corporation, according to a report by Tom Davenport and Jill Dyché for SAS Institute.

Although GE has had considerable success in recruiting data scientists, it is also creating an internally developed training program for them. It also occasionally has challenges recruiting data scientists who are familiar with the specific data issues around its industrial products, e.g., turbine sensor data, they said.

Hortonworks For Windows

Meanwhile, Hortonworks offers training for Windows developers. Called “Developing Apache Hadoop Applications for Windows,” it has both public and onsite availability.

Students will learn to develop applications and analyze big data stored in Apache Hadoop running on Microsoft MapReduce framework, as well as learn how to develop applications on Hadoop® using tools like C#, PigTM, HiveTM, HCatalog, Sqoop, Oozie and Microsoft Excel in this four-day course. Hortonworks said students should have programming experience, preferably with Visual Studio and SQL, as well as familiarity with the Windows Server operating system. No prior Hadoop knowledge is required.

Its target audience is .NET Developers and Data Analysts responsible for developing applications and performing analysis on big data using the Hortonworks Data Platform for Windows.

資料來源：Forbes