Machine Learning APIs As Provider Of New Information
Trust that adding new regions and Machine learning APIs will help your cloud compete Google Cloud Platform gets its first data center on the West Coast, after managing a data center for its own services in Oregon for more than a decade, Google Cloud Platform (GCP) has launched its cloud platform in the region. The data center of the new US-West is expected to reduce latency across the West Coast.In addition, Google has launched two automatic learning APIs for open beta cloud.
The company said in a blog post that the Oregon Cloud Region offers Google Compute Engine, Google Cloud Storage and Google Container Engine services, with two separate Compute Engine zones for high availability applications.Initial testing showed a 30-80 percent latency reduction for users in cities like Vancouver, Seattle, San Francisco and Los Angeles compared to Google Cloud “us-central” in Iowa.
Citing the video game industry as a key beneficiary of this reduction in lag, Google cited Paul Manuel, director of Multiplay Game Services, said: “Regional latency is an important factor in the gaming experience, the Google Cloud Platform Is one of the best we have worked not only from a technical point of view, but also in the one-on-one support we have received from the team. ”
GCP now consists of five regions – US West, US Center, US East, Western Europe and East Asia.
The company plans to add Tokyo as a region by the end of this year and says it will announce “more than 10” additional regions in 2017, an improvement over previous ratings that left the figure in ten.
“In order to implement fault-tolerant applications with high availability, Google recommends deploying applications across multiple zones and regions.
“This helps protect against unexpected component failures, including those in a single zone or region.”
Anxious to compete with Amazon Web Services and Microsoft Azure, Google has moved on to artificial intelligence to improve its own cloud.
Yesterday, it was revealed that Google’s DeepMind had been used to lower the PUE in some of its data centers by 15 percent. The company will now bring two IBM Watson-esque Machine Learning APIs into open beta.
The new Google Cloud Natural Language API (Machine Learning APIs) reveals the structure and meaning of the text in English, Spanish and Japanese (along with other languages in preparation). It says being able to “understand the general feeling of a block of text,” identifying relevant meaning in a block of text and “tagging them according to person, organization, location, events, products and media” in addition to identifying parts of the speech and “creating dependency analysis trees for each sentence in order to reveal the structure and meaning of the text.”
The company says that the Machine learning APIs could be useful for e-marketing because it allows analyzing the critiques of the product online, and enables the customer service centers to determine the “feel” of transcribed calls. The British food retailing company Ocado was mentioned as an alpha client.
In a conversation with the Wall Street Journal, the company revealed that it has worked with a large client to analyze more than two million minutes of customer service calls to determine whether the customer was satisfied or not.
“That has been a problem without a solution,” said Rob Craft, a Google automated learning product manager.
The Cloud Speech API, which offers businesses and developers voice-to-text conversion in more than 80 languages, is also open beta for both IoT applications and devices.
Google’s auto-learning products are very interesting, “said John Rymer, an analyst at Forrester Research,” but, “he added,” Google is not alone … Amazon and Microsoft still have a lot of lead. “