IBM turns to crowdsourcing for text translation project

Stephen Withers
Tuesday, 24 November 2009 04:53

IT Industry - Development

An internal IBM application that translates between English and 11 other languages embodies contributions from the company's 400,000 employees.

IBM's n.Fluent software is currently used by the company's employees to translate electronic documents - including web pages and instant messages - between English and Arabic, Chinese (Simplified and Traditional), French, German, Italian, Japanese, Korean, Portuguese, Russian, and Spanish.

With nearly 400,000 employees in 170 countries, IBM is well placed to tap people's ability to provide and update word translation pairs, with a particular emphasis on business and technical vocabularies.

IBM volunteers provided around 1.3 million words over a two-week period earlier this year, and there are current efforts to expand and improve the language data.

IBM envisages making the technology available as a product or service. A particular attraction of taking the product route is that - unlike web-based translation tools - it would provide greater security as there is no need for the text to travel outside the corporate firewall.

"To become a smarter planet, the world needs a shared vocabulary for collaboration -- particularly the business community," said David Lubensky, an IBM Researcher managing the n.Fluent project.

"We see n.Fluent as just such a tool, helping to expand commerce, cement relationships and make the world that much smaller, one word at a time," he added.

Other crowdsourced projects underway at IBM are aimed at making it easier for people to locate, transcribe, synchronise, tag, caption and share audio from video files.

The company identified real-time language translation as an emerging business area in 2006.