Google will introduce machine learning to catch bad apps

Google will introduce machine learning to catch bad apps

Android apps are of millions in number. Due to its vastness and open nature, security has always been a pain in the neck for Google. For the same, they have been criticized by many.

However, It seems that Google has decided to put an end to all the acquisition with a brilliant idea.

They will soon introduce machine learning and AI to trace out Malwares and exploiters in Google Play Store for good.

The technique Google is about to introduce has been named as ‘Peer Group Analysis‘. With this, Google will be able to scan and catch suspicious apps inside the Play Store.

Apps in the play store are usually categorized by metrics like purpose, description, metadata, statistics, etc. so that they form a peer.

From now on if a new security submission comes to Google for approval. It will be compared with its peer (Apps which has fulfilled laws of Google security) to find out if there is any unnecessary requests or permissions mentioned in it.

For instance, if an app is made for location tracking there is no need for it to access a users phone gallery. These kind of permissions are a serious threat to the security of billions of Android users out there.

In an official press release, head of Google security Martin Pelikan said that. The entire security team is focusing on apps which are a serious threat to the users and is positive about the outcome of the new implementation.

He even cited an example by saying that a flash light app does not need to access address book of a phone.

The news has been welcomed by Android users all over the world as people are unanimous about the need for advanced security. Particularly at this juncture of time.

In 2015 around 0.15% of total Android users across the world were affected by harmful apps. In 2016 it reduced to 0.05%, However, considering the 2 billion Android users the numbers are not small at all.

With the introduction of this new technique, Google is hoping to put an end to security threats to a greater extent.



Cost Calc.