Google used rival's database 'inadvertently'

IDG News Service - After evading the question for four days, Google Inc. folded on Monday and admitted in a blog post that developers at Google China had copied part of a software tool from rival Chinese Internet company Sohu.com Inc. for one of its own products. But Google later said the database was used unintentionally.

Google has yet to explain exactly how portions of a dictionary of Chinese words and names developed by Sohu -- which had not been made public or licensed for use outside Sohu -- ended up inside its Google Pinyin Input Method Editor (IME), saying only that it was an accident and the Sohu database was used to develop Google's product.

"Shortly after the product was released, we learned that content from a non-Google database had been inadvertently integrated into our dictionary," Google said Monday in an e-mail response to questions. The statement offered no further details of how the dictionary became integrated with Google's software.

On the surface, using part of a rival's copyrighted software in this way appears to violate Google's Code of Conduct.

"We respect our competitors and, above all else, believe in fair play in all circumstances; we would no sooner use a competitor's confidential information to our advantage than we would wish them to use ours," the Code says. "If an opportunity arises to take advantage of competitors' confidential information, remember: don't be evil. We compete, but we don't cheat."

While Google employees found to violate the Code "will be subject to disciplinary action, up to and including termination of employment," the company has not said whether such steps have been taken in this case.

On Sunday, Google released an updated version of its Pinyin IME with a new dictionary. That revision and an apology issued on Monday may have headed off a legal showdown with Sohu, but the damage to Google's reputation among Chinese Internet users was already done.

"Their image of innovation and 'don't be evil' was almost destroyed," said Jason Yin, managing director of market research firm In-Stat China, calling the events that unfolded over the weekend a "PR disaster" for Google China.

Pinyin IMEs are widely used in China as a way to type Chinese characters using their Pinyin romanization equivalents. Each IME draws on a built-in dictionary of Chinese words and names to suggest possible matches for users as they type Pinyin. These dictionaries take time and effort to compile, and ultimately determine the difference between a good IME and a bad one.

In the case of Sohu, two engineers spent more than a year compiling its dictionary, drawing on a database of popular search queries from the company's Sogou search engine.

Juan Carlos Perez, in Miami, contributed to this report.

Reprinted with permission from

IDG.net
Story copyright 2009 International Data Group. All rights reserved.