IBM, Sybase update their analytics packages with better unstructured data querying
Both IBM and Sybase have updated their lines of analytics software, optimistic of the growing market for such tools.
"We see business analytics and information management as a strategic growth area," said Bernie Spang, director of IBM's strategy and marketing for information management. The company expects to generate US$16 billion of revenue by 2015, just from work in analytics and information management, he said.
Sybase, which SAP recently began the process of acquiring, has updated its Sybase IQ column-oriented database management software, so that it can execute queries against unstructured data, that is to say the e-mails, Web pages, word processing documents or other formats that have not been indexed in some systematic way.
About 70 percent of corporate information is unstructured, said Tom Traubitz, the director of Sybase IQ product marketing, who quoted an estimate from IT analyst firm Forrester. By making this information searchable as well, organizations can get a wider grasp of potential trends and trouble spots, he said.
Sybase IQ, which is widely used in data warehouses, has a number of other new features as well, Traubitz said. It can support federation search, that is to say it can search for information across multiple databases, even other non-Sybase relational databases, such as Microsoft's SQL Server and IBM's DB2.
This version of the software has a range of new application programmer interfaces to allow programs written in new Web-oriented languages, such as Perl, Python, PHP and others, to access Sybase IQ directly.
The software will be available in June, Traubitz said.
IBM has also been beefing up its capabilities in analytics. The company has announced that it plans to incorporate into its analytics service offerings the Hadoop, an open source framework for breaking a search across multiple nodes and reassembling the results.
The IBM package of Hadoop will be called InfoSphere BigInsights, and like Sybase's offering, it too can use unstructured as well as structured data as source material. Hadoop "is an approach for gathering information from across a wider spectrum of structured and unstructured information, such as across the Internet or intranet, and doing analysis against it," Spang said.
IBM also upgraded the analytics software it acquired in its purchase of SPSS last year. The newly released version 6 of the SPSS Decision Manager predictive analytics software is the first to feature a Web browser-based interface, in addition to the regular thick-client version of the software.
The new interface "makes it very easy for the business user to take advantage of Decision Manager's model or to customize or build their own models," Spang said.
In addition to these updates, IBM has also developed a package for organizations, called SQL Skin for Sybase ASE, that wish to migrate their data and stored procedures off a Sybase ASE database, and onto IBM's DB2 database. This software is a follow-up release to last year's IBM package that allowed organizations to move their material from Oracle databases onto DB2 ones.