subject: Enterprise Search vs. Internet Search: The Paradox of Information Classification by:Jeremy Bentley [print this page] Imagine going into a travel agents and saying, "holiday Spain". What do you imagine you'd get from that? An armload of brochures to wade through and sent on your way? Or some helpful person might ask you some questions to determine what sort of holiday in Spain you're after.
Thanks to Google, this 2-word search mechanism has become widely used and understood. The millions of results you'd get back from Googling "holiday Spain" don't really give you the information you wanted, just as a pile of travel brochures doesn't help you narrow down your search for your perfect Spanish break quickly.
Using search on public websites consumers are happy, we're even trained, to begin the search for information like this, and accordingly our expectations of getting the right answer, straight away, is very low.
Despite our "quality requirements" being so low in internet searching, under the bonnet of all public-facing websites is considerable effort to deliver the right answer in response to the right search query. Websites are tagged with meta data; we submit content to directories and indexes; we spend money developing search engine optimisation and pay-per-click campaigns - all to try to understand what the searcher really wants, and get him to our content quickly.
Now consider enterprise search and information management. Anna in Accounts really needs the approved minutes from last Tuesday's budget meeting. Her quality requirement is very high. She doesn't want a list of all the minutes ever taken for budgetary meetings, or all the draft versions, she wants this particular document, and expects to find it at the top of the search results on page 1 of her company's intranet .
Here's the paradox: the effort that goes into correct meta tagging of content on the internet is extremely high, even though the users' "quality requirement" is quite low. But in the enterprise search, there is little or no effort put into the correct and sustained tagging and labelling of information, even though the users' quality requirement is very high.
Companies are just beginning to realise this, and the effect that not having a robust and efficient information classification system in place has on productivity.
Organisations now recognise the need for automatic content classification software, such as Smartlogic's Semaphore, which works with the existing content management system, Sharepoint, or Google Search Appliance, to automatically tag content according the company's agreed standards, making enterprise search and information management effortless and effective.
More information about Smartlogic's enterprise search and content classification software can be found at www.smartlogic.com.
About the author
Jeremy Bentley is the CEO of Smartlogic, the UK-based creators of Semaphore, a semantic platform that adds advanced semantic search capabilities to enterprise search and information management systems.
Adopting a semantic approach to information management delivers findability, improves the accuracy and efficacy of applying metadata to information and enables content integration across disparate sources.
Clients include NASA, The National Health Service, ABN Amro, Bank of America and The Office of Public Sector Information.
