Till recently people lean to the assumption that computers can’t understand humans. we’re complicated and in order for computer to understand us, they’ll have to know what feelings are. So far computer languages have done a great job telling the machine what to do and that was the reason why they were invented. However, further research to impend the human language in the machine to understand, were surrounded by uncertainty. That’s because computers couldn’t really capture the essence the human language, or so we thought.
Facebook announced that computers now can understand text based on deep text. It’s using a neural network for understanding what we write with a close to human accuracy. These networks can do that several times in one second and support more than 20 languages.
Human Interactions With Computers:
We regularly interact with computers thanks to automatic learning tools which understand humans more efficiently now. Siri and Cortana are a great example for that. It’s the same when asking google, the answers are for sure close to being very accurate with most of the related statistics.
Engineers at Facebook consider deep text to be an art because understanding the human language is accompanied with interchangeable feelings, places and daily events.
The deep text was programmed to monitor how people communicate and live.
It’s using a group of data which targets shares and comments as well as understanding feelings and words even those which are misspelled. All of that in order to teach the machine and train the neural network.
Uses Of Deep Text:
- It’s used for understanding comments. Comments are one of the closest ways to communicate with people thus filtering racist comments for example.
- It provides a better Facebook experience through understanding each user’s interests based on millions of pages.
- It give you suggestions whether you want a taxi or not.
- Further research will be done by Facebook to teach the machine in order to add more options to deep text.
For a video, please click here.