Google will expand the functionality of its enterprise Search Appliance, allowing corporate users to tweak the results of internal searches by giving added weight to content from specific sources or formats.
A new software update to the Google Search Appliance, launched in 2002, will allow for source biasing,enabling administrators to give search priority to content on certain servers or in document formats such as .pdf.
In addition, the software update will include results hit clusteringfor search results, with the search appliance grouping results into suggested topics. For example, if an employee searches for "customer" on the company network, a set of categories such as "customer support" or "customer contacts" will guide the search. Administrators would be able to customise the clustered results, Google said.
The Search Appliance would create the clusters on the fly, by using a statistics algorithm, with no major configuration by administrators needed, head of products for Google Enterprise, Matthew Glotzbach, said.
"We've been working on this for quite awhile, and we think we've gotten it right," he said.
Both of the major new features were designed to be easy to configure, he said.
"Other vendors have offered this type of functionality," Glotzbach said. "We make it very simple. We provide more of a menu-driven interface in a Web-based admin console, as opposed to scripts and writing code and tuning algorithms."
The new features would be available to Search Appliance customers as a free update within a few weeks, Google said.
The features weren't groundbreaking, but they should help Google compete in the large-business search market, senior analyst with Forrester Research, Matt Brown, said.
Google's Mini Search Appliance for small to midsize businesses was popular, but the clustering feature had been typically available in enterprise search technology about twice the price of the Search Appliance, Brown said.
The source-biasing feature took a different approach than most other search products, with its focus on source repositories instead of biasing by author or title, he said. Many large businesses could find Google's approach useful, and Google had been criticised in the past for not allowing administrators to tweak results.
"It just shows that Google is going to continue to move their products up market," Brown said.