October 22, 2012, 3:50 PM — At IBM's Information On Demand and Business Analytics Forum, being held this week in Las Vegas, the company announced a number of new add-ons and services designed to help organizations analyze their expanding data sets more quickly.
The new releases "are all around helping customers progress in their big data challenges," said Nancy Kopp, IBM's chief of big data strategy and marketing. "We want to help customers use all data types."
While companies such as Cloudera and Hortonworks may tout their enterprise Apache Hadoop distributions, IBM has taken a broader approach to big data analysis. Large organizations will want to consolidate their multiple tiers of data management into a single architecture, so the data can be shared across systems.
"There is definitely a play for a Hadoop system to make very large data sets," said Phil Francisco, IBM vice president of big data product management. But he noted that organizations will also want to make decisions on streaming data before it is stored on disk. And organizations still use their data warehouses to provide detailed analysis.
"The key is to have these [approaches] coordinated with one another, with information integration and information governance," Francisco said.
The new products and services IBM announced Monday help in this regard, Kopp said.
For its InfoSphere Streams real-time data analysis software, IBM has a set of pre-built report templates, called accelerators, that could help telecommunications firms more easily recognize common issues, such as fraud and customer churn.
Sprint is already using these report templates to monitor network events, outages and client use. Sprint "wants to analyze this stuff when it is happening," Kopp said.
IBM's in-house Hadoop distribution, called IBM InfoSphere BigInsights, has been augmented with new capabilities as well. IBM has generated new report templates, ones that can conduct sentiment analysis on data from social networks, such as Facebook or Twitter.
"The more accelerators we can build on top of our existing capabilities, the faster we can move our customers out of planning and into projects," Kopp said.
BigInsights now includes the federated search capability from the Vivisimo search engine, which IBM acquired in April. Using the Vivisimo interface, now called InfoSphere Data Explorer, users can execute a single search across multiple data repositories, including both structured and unstructured data.
"We no longer have these monolithic systems anymore. Your data is going to be in different workload-optimized systems. Having a federated search capability will be important to build these systems," Kopp said.