Web search goes smart

US technology giant IBM has unveiled an advanced search mechanism capable of extracting minute data from among billions of web pages that it says could become a crucial new revenue source for the firm.

WebFountain is claimed to auto-structure vast data into something meaningful

The company rolled out the new platform, dubbed WebFountain, while announcing a major deal with top content provider Factiva to exploit and develop the text tracking and retrieval system that it says is revolutionary.

  

“While a regular search just does key word matching, our platform is able to put context into the text (on the Internet) and customise it for a corporation or a user,” said IBM WebFountain President Robert Carlson.

  

“The challenge of doing this on billions of pages is a significantly different technical problem compared to what currently exists on the marketplace,” he said.

 

Powerful

  

“While a search (engine) is a very powerful tool, we think this is the next generation of (platforms) that mine large bodies of unstructured information and get the meaning out so that it can be used in business,” Carlson added.

 

IBM says the platform, the result of four years work at its Almaden research base in California, is able intelligently to use a massive data index that it builds up.

  

The system, run by a supercomputer that absorbs 25 million web pages a day from the internet, learns to recognise and put into context particular phrases and groups of words on command.

 

“The challenge of doing this on billions of pages is a significantly different technical problem compared to what currently exists on the marketplace”

Robert Carlson
president, IBM WebFountain

It begins to associate names and people in articles about a certain subject and gathers specific information that can be used as a business intelligence tool, he said.

  

The platform could allow companies such as advertising agencies and content providers to search huge quantities of text, and come up with the precise data they are looking for, saving critical time and manpower.

  

It would allow firms to track business trends or pick up scuttlebutt about their industry, themselves, their clients and their competitors by paying a fee for the use of the computer infrastructure each time they need data.

  

IBM believes that the platform could prove very lucrative for the company by helping to expand the market from an estimated $400-$500 million a year to between $5-$8 billion in five years, as customers become reliant on the system.

  

The company declined to say how much the system would cost to develop.

  

But while saying that there are many potential applications for the WebFountain platform, Carlson conceded that the system could run into initial problems, as text analysis skills are not yet as sophisticated as they could be.

Source: AFP