Google have made available a new major update to one of their most interesting, at least from a computer science perspective, APIs. The Prediction API allows developers to build smarter applications by leveraging Google’s machine learning algorithms to create many kinds of predictions.
The prediction API could be used to build recommendations, provide relevancy data, customisations, sentiment analysis, and many more features. All of this is available through an extremely compact RESTful API. The hard work with this technology is transforming your problem data into data that the API can work with and process effectively. There is also simple integration options with App Engine and Google Docs Script, as well as a set of client libraries to help out. The Prediction API builds on Google’s currently limited-release Storage API for recording the training data used in the algorithms.
New in this release is integration with the Google Code API console, which makes it much easier to work with the service and upgrade to the new paid tier for users with a higher than average demand on the system. The interface is also simpler, with automatic selection between algorithms and data type detection.
Find out more about the Prediction API on the Google Code Blog, or sign up for the API on the Code Console.
Comments