Google Brings Cross-Language Translation to Search Appliance
Many enterprise search vendors offer their search software in multiple languages to help workers all over the world sift through corporate documents in sales, human resources and other areas.
Google Enterprise Labs, an effort Google launched in 2007 to test potential enhancements to its enterprise search products, has taken the polylingual concept up a notch by providing a Cross-Language Search feature in its Google Search Appliance.
The feature instantly translates search queries for internal company documents, sent from users' PCs to the GSA, in any of 34 languages, Cyrus Mistry, Google Enterprise product manager and Enterprise Labs creator, told eWEEK.
"This is analogous to giving every employee in a business 34 translators sitting at their desk and translating everything they want to look for within a 10th of a second," Mistry explained. "It would take a massive investment for companies to have translation servers on-site."
For example, if you're an English-speaking employee and you want to find and translate a document written in French from an office in Paris, you can do so with a few simple keystrokes. Your PC sends the request to the GSA, now imbued with Google's machine translation software, which does the work and renders the file in English on the fly.
Conversely, this feature will help those whose native language isn't English find and translate documents from U.S.-based employers into their native tongues.
Cross-Language Search, which can be downloaded here, is the latest of 10 features to roll out from Google Enterprise Labs since its inception in October 2007.
The most popular is the search-as-you-type feature, which presents suggestions and auto-completes queries before the user finishes typing. Another, do-it-yourself key-match lets users promote useful results to help colleagues in their searches.
Enterprise Labs works like other labs efforts at Google, including Gmail Labs and Google Apps Labs. That is, Google's 10,000 or so engineers build an experimental feature and launch it within Google for testing.
Once Mistry collects the feedback, he "sanitizes" the feature and releases it into the general public via Enterprise Labs. If he likes what he sees based on customer use, he rolls it into the product it's designed for, which is most often the GSA.
Cross-Language Search highlights what could be a painful reality for rival enterprise search providers, including Microsoft's Fast unit, Vivisimo, Endeca and Autonomy.
While Mistry readily acknowledges that rival platforms offer enterprise search in multiple languages, these much smaller vendors don't have machine translation experts to create such features, he said. Google has the benefit of having these experts for its consumer search offerings.