Google I/O last week left us with Google TV and IT-oriented Android OS news, but it was the release of the new APIs that interested me the most. Google clearly has its black belt in API development, and the current bunch of APIs continues that trend. Most interesting was the new Prediction API, which provides advanced analytic services for any Internet-connected applications, on premise or in the cloud.
Google says "the Prediction API allows you to get more from your data and makes its patterns more accessible. Specifically, the Prediction API leverages Google's machine-learning infrastructure to give you the tools to better analyze your data and reveal patterns that are often difficult to manually discover. The API also enables you to use those patterns to predict new outcomes, which facilitates the development of all types of software, from textual analysis systems to recommendation systems."
[ Get the no-nonsense explanations and advice you need to take real advantage of cloud computing in the InfoWorld editors' 21-page Cloud Computing Deep Dive PDF special report, featuring an exclusive excerpt from David Linthicum's new book on cloud architecture. | Stay up on the cloud with InfoWorld's Cloud Computing Report newsletter. ]
You implement this API using a three-step process. First, upload the training data into Google Storage (Google's cloud storage service). Second, build the model from the data. Finally, make new predictions using that data -- for example:
POST www.googleapis.com/prediction/v1/query/${mybucket}%2F${mydata}
For now, you can use the Prediction API for free, since Google considers it to be in beta. Access to the Prediction API is currently by invitation only. To request access you need to sign up on the Google waitlist. (I did so before writing this blog post.)
Predictive analytics is not new; those in the BI world have been doing things like this for years. However, that's with some very expensive hardware and software; thus, it's been out of reach of many smaller enterprises and certainly impractical for most common application development efforts.