It is the use of type of multi layered ML model based on neural network various layers can be automatically determine how best to use input data to make predictions through process called training.
Deep Learning as primarily been focused on computer vision application today , problems like identifying cats on repository of images.
Researches are just starting to applying DL to other types of problems like Natural Lang processing, understanding languages as used typically used by humans.
Facebook announced the new text understanding engine that called DeepText. Deep Text is used by Facebook internally, not for open sourcing (as of now). It is developed based on research papers.
By applying DL to the problem of Natural Language processing Facebook is able to overcome number of challenges they faced with traditional NLP techniques. They found that DeepText is easier for them to scale, less dependent on the specific individual languages and more efficient way to process the data available to them.
In Facebook blog post they mentioned that DeepText engine can understand the text content what they called “near Human accuracy”. They read the several thousand FB post per sec across over 20 languages.
It is built top of FBLearner and this is AI platform. DeepText team is focused on application in the area of conversation understanding.
Text understanding includes multiple tasks, such as general classification to determine what a post is about — basketball, for example — and recognition of entities, like the names of players, stats from a game, and other meaningful information. But to get closer to how humans understand text, we need to teach the computer to understand things like slang and word-sense disambiguation. As an example, if someone says, “I like blackberry,” does that mean the fruit or the device?
Text understanding on Facebook requires solving tricky scaling and language challenges where traditional NLP techniques are not effective. Using deep learning, we are able to understand text better across multiple languages and use labeled data much more efficiently than traditional NLP techniques. DeepText has built on and extended ideas in deep learning that were originally developed in papers by Ronan Collobert and Yann LeCun from Facebook AI Research.