Menu

Instagram : from machine learning in order to make the network more enjoyable

Instagram deploys two new tools on its social network in order to make it better. The offensive messages and spam should be reduced with the help of machine learning.

If the essence ofInstagram is the beauty through the picture, the social network is often rotting by the comments found there. Between hate messages and spam, it is sometimes horrible to take a look in the jungle of the responses can be found in a photo. Instagram is aware of this and finally decides to make moderation a bit more effective through the use of two tools to avoid everyone to do so and therefore to be confronted with unpleasant messages.

The comments inappropriate

The messages “toxic” can now be automatically filtered out. Thanks to this, all the comments consisting of “words or phrases often deferred as inappropriate ” can be automatically hidden. A filter manual also allows you to choose words or expressions that will then be used to filtering.

To activate, go to your account settings and then drag to the sub-menu ” Comments “.

The other comments always appear in the same way and it is always possible to moderate on a case by case deleting them or posting them.

The learning machine against spam

An anti-spam filter has also been set up to remove thanks to the learning machine messages too in English, Spanish, Portuguese, Arabic, French, German, Russian, japanese or chinese. What block automatically to a large number of spam comments.

Very humble, Instagram, admits that ” the work is still far from being finished or perfect “, but it is good to see that the network is increasing day by day, either by adding new features — often inspired by Snapchat –, in enhancing opportunities or by optimizing the application itself. Keep in mind that the comments are not the only elements of this social network to have a bad effect on the mental health of young…

Instagram

No comments

Brought to you by:

Your AD Here ! Contact us at admin@appnews.be

Subscribe to Blog via Email

Enter your email address to subscribe to AppNews and receive notifications of new posts by email.

Official Fan Page