Small business wins with big data
- 01 May, 2012 10:26
A recent US CIO article on the burgeoning phenomenon of Big Data, published in Reseller News, made much of the need for modern semantic search engines. There is a New Zealand-developed option, SYL Semantics, which is attracting overwhelming interest from both government and commercial interests, according to chief executive Sean Wilson.
"We're also talking to several multinationals which are interested in our technology," he says.
SYL has signed up three government departments so far, but Wilson says he is not allowed to name them at this stage. That's not bad going for a company which was launched less than a year ago.
Around 90 percent of all data was created in the past two years, and it's estimated that users spend up to 25 percent of their time searching for the appropriate document. Hence the term Big Data.
Semantic search improves search accuracy by understanding the contextural meaning of terms. The goal is to deliver quickly the information sought, rather than the user having to search through a long list of loosely related keyword results.
The concept is not new. There are several semantic search engines, all with varying degrees of focus and success, according to technical reviewers.
SYL has a New Zealand patent for its technology and is applying for a US patent. Wilson describes SYL Enterprise Search as the next generation of enterprise search applications.
"It uses semantic technology to create a smarter search experience by understanding the meaning of what the user is looking for and using similar or related search terms to identify the relevant documents. The system learns over time to provide more relevance in less time. We call this intelligent enterprise findability."
Custom technology is supported, including dictionaries and industry terms (such as the acronyms used by the military and police). Content can be federated across content repository boundaries such as the web, intranet, email or a CRM system for single enterprise search results.
The company has just launched Version 1.8, which chief technology officer Adrian John says builds on an earlier version at a government agency. "Version 1.8 addresses aspects of scalability and provides a significant improvement in performance," he says.
"We can now cater for security models in bespoke applications. When an organisation builds its own data management tool, there will often be a customised security model for that application. SYL can now harvest that data."
The new version also allows content to be fed into the system using XML files. It also allows content to be ingested via web services.
"The ability to import custom security is a real differentiator for us," John says. "A lot of search engines can't deal with that."
He says that so far the technology has been delivered as an appliance but will soon be able to be fully embedded in a solution.
"We're talking to a large integrator about putting SYL into the cloud as we head to Version 2.0."There is a programme of work scheduled for the next six to eight months to deliver the new version."We are continuing to concentrate on scalability and enhanced features of functionality to natural language processing," Wilson says.
"The key is that SYL does things the enterprise expects. It is smart, practical and highly geared toward the enterprise."
He says enterprise search is valuable to any organisation, regardless of size. "It's about searching over all of your content respositories. SYL is a productivity tool, and it also saves on a tremendous amount of frustration for users.