HP Breaks Autonomy IDOL Into Discrete Services

IDOL functions such as sentiment detection and entity extraction will be available as discrete services

Developers seeking to embed some advanced text processing capabilities into their applications might want to take a look at a beta service that the Autonomy software unit of Hewlett-Packard is fielding, tentatively called IDOLOnDemand.

The platform as a service (PaaS) will, over time, offer most all of the functionality of HP Autonomy's IDOL (Intelligent Data Operating Layer) search and content processing platform, which had been available only in an enterprise software package.

"If we want to be successful as a platform today, we have to do more than create a large installable product. We have to enable it so developers can use it," said Robert Youngjohns, senior vice president and general manager of HP Autonomy.

The service will expose most of the IDOL features as discrete services, accessible through APIs (application programming interfaces). HP is hoping that enterprise developers use the service to embed IDOL functionality into their own applications.

HP announced the initiative in December and plans to formally launch the service in April or May.

In preview mode, the company already offers a number of services on a trial basis, such as barcode reading, language detection, format conversion, entity extraction, face detection, text extraction from images, and sentiment analysis.

Over the course of the next few months, HP will add more functions, such as geolocation lookup and API discovery. IDOL has about 400 functions centered around the searching, filtering and packaging of unstructured data.

"Rather than have all these functions in a homogenous product, why not expose them individually as Web services?" Youngjohns said.

The PaaS approach provides an easy way for a developer to build a software program without writing all the components from scratch, or importing third-party libraries.

For instance, a developer trying to parse a large amount of customer feedback could have the IDOL sentiment analysis function analyze the feedback and return an estimation of how many comments were positive and how many were negative, freeing managers from having to read through all the comments themselves.

The IDOL services will fall into two broad categories, stateless APIs and those that maintain a persistent state within the HP Cloud, where the services are run.

HP is first working on building the stateless APIs, where the service request along with the required data is submitted and the results are returned.

Over time, the company will also offer stateful services, where users can store their data in an object store hosted by HP. The current version of IDOL, the recently released IDOL 10.5, will serve as the basis for the service offering.

"Sometime in the future, we will have all of our services in IDOL exposed this way," Youngjohns said. The company will also introduce a policy manager that will work with these services.

"A lot of what we do is about how you apply policy to data," Youngjohns said.

HP competitor IBM has a similar idea. This week, IBM announced that it would be offering much of its middleware software portfolio as cloud services.

HP has not decided yet on how pricing will work with these services.

Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson. Joab's e-mail address is Joab_Jackson@idg.com

Recommended
Join the discussion
Be the first to comment on this article. Our Commenting Policies