GLOBAL - IBM envisions business mash-ups with 'just a few a few clicks'.

Workers will soon be able to put together a mash-up using enterprise data with the same ease that it takes to create a web 2.0 application, IBM predicts.

Mash-ups are services that combine information from multiple sources to create a new service.

Popular examples include online maps that allow users to find listings for rental apartments and homes for sale, or even available shelters after a disaster.

"This idea of the Long Tail is applicable not just to data, but to applications," Rod Smith, research fellow and vice president of emerging internet technologies at IBM, told www.vnunet.com. 

'The Long Tail' describes a phenomenon whereby a large number of blogs or users can create a lucrative niche market for goods or advertisers. 

The term was coined by Chris Anderson, editor-in-chief of Wired magazine, and is considered one of the driving forces behind web 2.0.

In business applications, the Long Tail could allow firms to cut back on research and development spending, enabling them to switch from a mass marketing to a niche marketing model.

Firms could switch from offering a few products to millions of people, to offering millions of products that cater to small groups of customers.

A business version of a web 2.0 application could allow companies to combine weather information, for example, with their enterprise data.

IBM has built a service that combines information about ships with weather data, allowing workers at a shipping company to track the location of a vessel, its route and destination.

To the end user, web 2.0 might look like it is all about combining data streams. But it requires a large IT infrastructure component to make it work for businesses.

The data stream needs to be made available as an XML feed to allow it to be coupled to other information, and often has to be pulled from legacy applications that lack standard support for such features.

The infrastructure aspect will be delivered by service oriented architectures, another emerging technology that has mostly focused on the reuse of code within a company.

But the technology also allows software to be built from ready-made components, each of which offers a unique functionality.

"Web 2.0 centres on two axes here: data and service oriented architectures," Smith said in an interview.

"SOA is what is going on under the covers. Some customers will say that mash-ups are the last mile for SOA."

IT departments, meanwhile, are in for a culture shock because they will lose their monopoly on developing new applications. In a web 2.0 world, workers will start to create their own tools without any oversight.

As a result, software will more closely meet user demands than in the current development model, again cutting back on overall development time. Currently user requests are often lost in the translation to the developers.

Although the emergence of the business user as an application developer was not planned, business users grew accustomed to this model when they started using blogs, wikis and other web 2.0 tools readily available on the web.

The same trend has also been cited as a reason behind the success of Salesforce.com

"Business users need [new services] now. IT in many cases cannot start projects for six months. They need it for a small period of time and they want to service to be disposable," said Smith.

"It is just the opposite of the requirements that IT has always been told to meet: build it once, build it resilient, build it to last forever. That is why the cost of doing these applications has always been so high."

IBM is offering early tools that allow for the creation of mash-ups such as its QEDWiki tool. But Smith expects to start rolling out more advanced software in the second half of this year.

Much of IBM's support for the open source Dojo framework that creates Asynchronous Java and XML developer tools is also focused on its plans for business web 2.0. IBM has committed four developers to the project, Smith said.

Even with the development tools in place, there remains much work to be done. Companies need to establish standards for securing their mash-ups and must determine how they honour copyrights when mixing data from themselves, partners and outside providers.

Although Smith is confident that mash-ups will become a mainstream technology faster than others such as SOAs and web services, he warned that it will still take several years to catch on.

Copyright © 2009 vnunet.com