Kokonutguy, I have that newspaper, because I am in Dubai right now. The gulf news published an even better article in their tabloid, about the Engines of change on the super-highway!! Neomedia fits right into the missing part! the story is on search engines and displaying information(type anything and you will get 500.000 results)! I was actually thinking of sending the author of the article an email and mention Neomedia!
This is one of the articles he wrote, there are so many! I don;t even have to tell you, you know! For all other users I have a whole document, everytime Neomedia fits in!! Again and again I say, Neomedia has unlimited options!!
The quest to find a better search engine by Richard Waters
The rise of the search engine as a tool of modern life reflects the huge increase in the amount of "unstructured" data that exists both on the world wide web and inside corporate information systems. Structured data, as the name suggests, resides in organised databases: laid out in tabular form, it is easy to abstract and analyse. Unstructured data, on the other hand, is a mess. Information in an e-mail or a Word document follows no organising principle. Yet this, increasingly, is where much of a company's most important information is to be found. FT.com /Management
The quest to find a better search engine
By Richard Waters
Published: May 11 2004 21:05 / Last Updated: May 11 2004 21:05
Physicists are just like the rest of us in at least one respect. When they go online to search for information, they expect the earth.
Bebo White, who runs the website for the Stanford Linear Accelerator Center, a high-energy physics laboratory in California, knows this all too well. "They get very frustrated if we don't return good results, and quickly," says Mr White.
And what do the expert visitors to the centre's website look for most often? That day's cafeteria menu, says Ruth McDunn, another technician on the site.
Call it the Google Effect. Expectations of search engines have skyrocketed. Whether it involves complex specialist knowledge or the completely trivial, there is a general belief that everything should be available instantly, at the click of a mouse.
While Google has turned this new desire into a business with spectacular effect, it is an expectation that has started to bedevil companies the world over. Customers, employees and regulators all now demand instant access to information.
It does not matter whether they are visiting a company's public website, trawling its intranet or trying to mine information from a mountain of employee e-mail: the high expectation of users is the same.
The problem of finding relevant information inside a large and complex company or government agency has created a wealth of new technical responses. The companies involved boast that their technologies are already far more sophisticated than that of Google, and represent the next wave in search.
"The whole search engine story is very far from being over," says Mike Lynch, chief executive of Autonomy, a British company that specialises in search technology. "The technologies that are being used are only a small part of what is possible."
The rise of the search engine as a tool of modern life reflects the huge increase in the amount of "unstructured" data that exists both on the world wide web and inside corporate information systems. Structured data, as the name suggests, resides in organised databases: laid out in tabular form, it is easy to abstract and analyse. Unstructured data, on the other hand, is a mess. Information in an e-mail or a Word document follows no organising principle. Yet this, increasingly, is where much of a company's most important information is to be found.
For "enterprise" search engines, the challenge is very different from the one faced by Google on the web. Google's PageRank technology tackles the problem of sifting the millions of web pages by analysing links between sites: the more links that lead to a particular page, the more useful its information is deemed to be and the higher it will appear in a search result. Google has since refined its search engine by overlaying other technologies, though these remain secret.
The value of assessing a page's popularity has its limits, though. "Google's technology is wonderful for the web at large, but it is not nearly as clear it is generically useful for companies," says Mr Andrews.
Looking at popularity "doesn't help much with the way we want to access information," adds Ms McDunn at Stanford. This has not stopped Google entering the enterprise search business, though, selling its software pre-installed inside a server known as a Google Search Appliance.
Another problem is that while Google can be excused for delivering only its best estimation of results drawn from the web at any one moment, a corporate search engine must meet high standards of accuracy. And the data may need to be abstracted from more than 200 different document formats - many more than the standard HTML, PDF and Word formats encountered on the web.
It is these problems which a gaggle of specialist search engine companies are now trying to tackle, says Mr Andrews. Much of the brainpower is being applied to write new algorithms that are better at assessing relevance.
According to Prabhakar Raghavan, chief technology officer at Verity, a US search company, most of the basic approaches to assessing relevance have been known for some time. The art now lies in finding ways to blend these different methods in ways that produce better results, a process known as "tuning".
To produce the next big leaps in search, technological innovation is now being applied in two main areas. One involves imposing a rudimentary structure on unstructured data through the use of automated classification. "When you take unstructured content and structure it, you add value to it - you expose it to an application that can exploit it," says Mr Raghavan.
While "taxonomy" has become one of the biggest buzzwords in search, though, the attempt to impose more order on the chaos remains at an early stage. The second new buzzword is "personalisation". Users are losing patience when having to sift search results to find information that suits them, say search engine companies.
Part of the problem lies in the way most people use these online tools. They usually fail to be specific in their search requests, says Ali Riaz, marketing director at FAST, a Norwegian search company.
"Human beings are very lazy about giving information, but they're very greedy about taking it," he says.
This is a characteristic of the physicists who visit the Stanford linear accelerator site. They seldom enter more than one word in the search engine, then get impatient if the results are not ranked in a way that closely matches their particular interests, says Mr White.
Guessing at what responses a user might want, depending on their job inside a company, is one approach that is already being used. Employees in marketing and legal affairs might reasonably be expected to want different types of information from a search, even when they enter the same search term, says Mr Raghavan.
Mass-market search engines like Google and Yahoo are also working to fit their results to the specific interests of users.
With so much work underway, most experts confidently predict that more big breakthroughs lie ahead - even if the predictions seem touched with an understandable dose of Google envy. "A new and better search engine will appear, and much sooner than anyone expects," says Mr Lynch at Autonomy.
News 
Market Data 
Discover 
AI
