[ad_1]
Meet Bert. Developed by Google alongside the Allen Institute for Artificial Intelligence, the AI system, Bert can complete the missing parts of sentences almost as well as some humans.
Thanks to his basic understanding of the language, Transformer's Bert or Bidirectional Encoder Representations can take a phrase like "the man went into a shop and bought a ____ of milk" and filled it in with appropriate words.
While this may sound like a simple task, this is not an easy task for AI and can pave the way for a wide range of possibilities in artificial intelligence.
via GIPHY
Natural language processing
For the uninitiated, NLP, or automatic language processing, is the cornerstone of the latest developments in artificial intelligence. NLP is a tool that allows computers to communicate effectively with humans, including different components such as speech recognition, sentiment analysis, and machine translation.
The sub-domain of artificial intelligence, Natural Language Processing, brings computers closer to an understanding of language at the human level. As you can imagine, computers do not intuitively understand how humans do it. "Understand" even the simplest sentences can be very difficult.
Because computers can not necessarily read "between lines," researchers use the latest advances in machine learning and deep learning to perform tasks such as linguistic translation, semantic comprehension, and text synthesis. You have probably already encountered products already using an earlier version of NLP.
The rise of IA
Bert from Google is an important leap forward in the world of Artificial Intelligence. Such a computer system can learn, understand, and handle small idiosyncrasies of language and use them to perform specific tasks. There could come a time when you could have a complete conversation with a computer system.
Specifically, Google's NLP program and the Allen Institute could be used to improve technologies that you already have at home, such as Alexa and Google Home, as well as to automate certain business practices.
Bert learns by analyzing millions of sentences in books of all kinds, including novels of love and science fiction. Through this vast library of data, Bert has learned to guess missing words anywhere in a sentence, in order to understand the basic relationship between words. This allows Bert to predict the missing sentences, whether it's words or even small stories.
With this same skill and his same knowledge, Bert gives explanations to the questions, or even measures the feeling of a film critic, responding if the criticism is positive or negative. Bert was even able to tackle the Allen Institutes Common Sens test.
Although Bert is a big step in the right direction, the researchers are not about to want to be with AI and Bert is still far from having the human good sense.
via GIPHY
Who knows, in the near future, you could have late night conversations with your home system. Bert is currently trained in 102 languages.
Via: Cornell University Library
[ad_2]
Source link