The search giant is reportedly said to be testing Gboard on Android first to improve the Google mobile keyboard‘s suggestions. This is something like Apple started testing different privacy, a method to gather behavior data while anonymizing users identities, which Google is now doing. Here are the details you will be interested in knowing about.
Last June, Apple, the Cupertino-based company expected the same would improve QuickType predictions, that Google jas just begun trying out a similar method with Gboard to improve its automatic suggestions, but has taken a different direction to ensuring privacy: approach in keeping data on the device, not uploading it to the cloud
How does it work? It made this happen by downloading the latest text prediction mode to your device, improves it by learning from behavior data on our phone and then sending a summary of the changes to the cloud This is actually be the combination o the their single-device updates and a new shared prediction mode which is created to download and start the process all over. While Googles’ research scientists are calling this method ‘Federal Learning’ by means.
Keeping the learning process local on your device by uploading small summaries to servers instead of large data batches reduces both power drain and bandwidth required. When compared to Apple’s technique, which adds “mathematical noise” to user data in order to protect identities, Google’s tech might make it less of a strain on devices and cloud services.
Google’s testing Federal Learning out first on Android’s keyboard, Gboard, to improve its word suggestions. In the upcoming days, or near future, it might be used to also improve each user’s own personal language models in native Gboard, as well as adjust photo rankings based on which types people look at, share and/or delete. Stay tuned for more!
You might also like to check out: