Information retrieval solutions that don`t incorporate innovative ways of handling organisations` rapidly growing datasets are soon going to be ineffective.
Today`s employees have far too many documents to read and, without the support of more advanced tools in their search solutions, relevance ranking becomes increasingly difficult, says Iain Fletcher, international sales director of Convera International in the UK, following a recent visit to SA.
During his stay in the country, Fletcher held successful discussions with several government and financial institutions in conjunction with Knowledge Focus, sole African distributor of Convera`s RetrievalWare.
"Most companies understand the value of information and how critical it is for informed decision-making," Fletcher continues. "Although many have deployed search technologies to unleash the wealth of mission-critical information locked in their datasets, these are becoming increasingly ineffective, owing to the rapid rate at which datasets are growing."
Fletcher points to the US Federal Bureau of Investigation (FBI), as well as major intelligence-gathering services in the US, UK and elsewhere have selected RetrievalWare as a search and categorisation platform within their investigative data warehouses. After 11 September, the FBI created a sophisticated Secure Collaborative Operational Prototype Environment (SCOPE) with a counter-terrorism and intelligence data repository. RetrievalWare was chosen as the repository to improve the sharing of intelligence information and collaboration across multiple government agencies, enhancing the US government`s ability to prevent terrorist attacks. The decision was based on RetrievalWare`s ability to cut through enormous amounts of data to find minute details required by agents to respond to possible homeland threats.
"Organisations such as the FBI are usually the first to encounter problems caused by burgeoning datasets, simply because they gather more information than any other organisation or business. Consequently, Convera technology was specifically developed for intelligence-gathering agencies, and is based on the principle that if the technology works for them, it will work for anybody. Whatever problems the FBI and similar organisations experience, others will experience two to five years later, but RetrievalWare will already provide the solution."
Fletcher says most search solutions are generic in nature: "Optimised for no one but able to be used by anyone. The search environment in RetrievalWare, on the other hand, has been `verticalised` to support most major industries.
"The search requirements of say, financial institutions differ from that of other industries," he explains. "People who want to find information on a specific industry will perform searches using words specific to that industry. In the finance industry, for example, there are more than 10 000 concepts which you won`t find anywhere else. If the search technology thoroughly understands this particular environment, better results will be achieved."
According to Fletcher, Convera plans to release up to 20 off-the-shelf, highly detailed and industry-specific taxonomies in 2004 that can be modified to suit any organisation. "Some have already been developed and are due to be released soon," he says.
Additionally, version 8 of RetrievalWare includes a brand-new Dynamic Classification module that makes the process of browsing documents highly personalisable for company departments and individuals. "This means everyone is not limited to the same fixed view of data," says Fletcher. "Whether they are a clerk in the accounts department, a marketing director or a senior engineer, they can examine the data in ways that are optimally useful to them.
"We are extremely excited about the new functionality in RetrievalWare 8, which constitutes a giant leap forward for search and retrieval technology. By providing a single integrated platform where users can personalise search results through the use of dynamically generated classifications, RetrievalWare 8 organises and delivers enterprise content in the context of the request, regardless of the format, language or storage location of the retrieved data.
"With dynamic classification, businesses can slice-and-dice search results, which allows users to discover hidden information by identifying previously unknown relationships among data entities," Fletcher concludes.
Share