Google claims great development in voice search business

Google Vice President of Search Business, Johanna Wright

Today, Google announced that with the application and improvement of new technologies, the company's voice search business is developing rapidly.

Last year, Google added a machine learning technology called "Deep Neural Networks (DNN)" to its voice search service. With the support of this technology, Google reduced the error rate of voice recognition by about 20%. Since then, Google has been continuously improving this service and helped increase the accuracy of Google's voice recognition by another 25% or so.

Today, Google held a launch event for its new flagship phone and operating system in San Francisco. At the event, Johanna Wright, Google's Vice President of Search Business, also announced the aforementioned progress in voice search. According to Wright, Google has long been committed to improving the quality of voice search. For the past two years, Google has focused on the rigor of voice recognition in the direction of voice search.

Wright claimed that voice search hasn't undergone drastic changes, but there have indeed been a series of improvements and optimizations, just like Google regularly improves its search engine algorithms.

Google’s voice search has been integrated into various Android operating systems and even further integrated into the newly launched Nexus 5 product. In addition, Google's voice search has become an effective application for iOS and Windows 8 operating systems. Furthermore, Google provides downloadable web services for its voice search application.

Of course, when describing the progress of its voice search business, Google still lacks absolutely accurate data, and the industry questions whether Google is exaggerating the progress of its voice search business. However, since voice search is increasingly becoming an important channel for controlling mobile phones, Google's progress in voice search makes sense.