Gateway to data heaven

While the 16th century philosopher Francis Bacon equated knowledge to power, today knowledge is as likely to be related to...

While the 16th century philosopher Francis Bacon equated knowledge to power, today knowledge is as likely to be related to revenue.

In fact, IDC has recently estimated that Fortune 500 companies are experiencing a $12bn 'knowledge deficit'. This is described as finances lost spent searching for relevant data or basing decisions on inadequate information.

And the problem will not go away. Here's another Carl Sagan-like statistic, this time from IBM, which states that more information has been produced in the last 30 years than during the previous 5 centuries. Currently, there is 1 exabyte (1 times 10 to the power of 18) of online data in the world, equivalent to telephone books stacked to the moon and back again. Apparently, this amount of information is doubling every 12 to 18 months.

Cue IBM's Enterprise Information Portal (EIP), a new offering that will, says IBM, allow companies to pull together valuable information from all sorts of sources, such as business intelligence, business applications, unstructured content, and structure data. But this is a 'portal' with a difference. Rather than being a community site or trading hub within the public domain of the internet, Mike Blake, data marketing consultant at IBM, refers to EIP as a portal service providing 'heavy infrastructure' for users within the virtual four walls of an organisation.

'There is all sorts of information that is of value to today's organisations,' says Blake. 'Competitor information, data from the internet, knowledge that is hidden deep in the company's databases. The job is to integrate it all.'

IBM is combining some of its cutting edge technologies on this project, such as the DB2 universal database, MQSeries, Websphere and its Content Manager offering. Big Blue regards EIP as a final building block in IBM's e-business (ie the paperless office) strategy for companies which are struggling to consolidate related information stored in different formats and different internal systems. It is a concept best explained by example.

Take a teleworker at an insurance company who is dealing with a customers claim following a recent car accident. He will need to see a copy of the customer's insurance policy, details of previous correspondence, perhaps pictures of the damaged vehicle, police reports the list goes on. All these pieces of information, vital to making an informed decision will be stored in different repositories.

Mike Blake explains the dilemma faced by a user as we creep towards the nirvana of a paperless office. 'He will say to himself 'I need this information, Where is it stored?, Can I access it?, Can I bring it together with my other information?'

The customer's policy details may be on a structured database, DB2, Oracle, or SQL Server for example. The records of previous correspondence will more than likely be on some kind of bespoke customer tracking system. Any digital photo images required will be found in an archive store while the police report may be scanned into another document repository.

For the archiving of these latter objects and for that matter video and sound files, IBM has re-badged some old DB2 technology into a single product now called Content Manager. This deals with the archiving and organisation of data not suited to the parameters of a relational database, often referred to as 'soft data' or blobs (binary language objects). It is relatively old technology that, with the surge in e-business and the need to make use of all available information to aid customer service and information sharing, has taken on a new relevance.

The EIP sits on a middle tier NT box above all these separate databases and information stores and pulls together all this information. IBM has written API connectors from EIP to Content Manager, DB2 and other competing databases, while MQSeries and what is described as 'advanced federated search' technology within EIP enables data integration and the pinpointing of pertinent information.

While this is all going on at the backend, the information will be published to the company intranet web browser via Websphere. To present the user with a graphical user interface, IBM has partnered with several portal companies such as Plumtree and Mediapps. To use a car analogy, this is the 'dashboard' on top of IBM's 'gears and engines' beneath the 'bonnet'.

Ian Wells, md of the UK subsidary of Mediapps is keen to extole the added value his product Net.portal offers, up and above simply publishing information requested from the bowels of a company.

'We are what they call in the US, an 'information aggregator' - that means as well as porting to Websphere and Domino, we also provide additional information from the internet through agreements with content providers and news wires. This will be, for example business info, news and statistics that is of benefit to the user.'

While no company in the UK has yet adopted EIP, since its launch in December 1999, Blake says IBM is talking to several FSTE 100 organisations.

Meanwhile, IBM will continue to include further enhancements to EIP.

Read more on Privacy and data protection