Artificial Intelligence Software, SyntaxNet, is now Open-Source
Google open-sourced its artificial intelligence software, SyntaxNet, aiming to get more human brains attacking the challenge of natural language understanding by computers. Ultimately the company hopes to accelerate the progress of natural language research and development by provided free access to their software.
At it’s core, SyntaxNet is basically a parser, which means, utilized properly, it can help you discover how words can relate to each other based in linguistic elements. In almost any type of NLP task, parsing is inevitable and is a hard challenge to solve. It is also computationally expensive, so having a freely accessible open sourced framework such as this will significantly accelerate innovation for academics and data scientists alike. In the case of SyntaxNet, the framework learns to understand sentences by analyzing millions of sentences. But these aren’t just any sentences. Humans have carefully labelled them, going through all the examples and carefully identifying the role that each word plays. After analyzing all these labeled sentences, the system can learn to identify similar characteristics in other sentences.
Google is also sharing a pre-built natural language processing service that it has already trained with the system. They call it, well, Parsey McParseface, and it’s trained for English, learning from a carefully labeled collection of old newswire stories. According to Google, Parsey McParseface is about 94 percent accurate in identifying how a word relates the rest of a sentence, a rate the company believes is close to the performance of a human (96 to 97 percent).
Ultimately, SyntaxNet as a stepping stone to much bigger things. Syntactic parsing merely provides the foundation. Many other technologies are needed to take the output of SyntaxNet to truly grasp the meaning of language. Google appears to be opening-sourcing the tool in part to encourage the community to look beyond syntactic parsing. And, both academia and the development industry stand to benefit from Google’s latest endeavors because because of it. The framework that they released is one of the fundamental elements needed to perform language understanding.
Google seems to be telling the artificial intelligence community to not worry about parsing; that got that covered and it’s now a given. Now data scientists can explore go harder.
Read more about Google’s efforts in the VentureBeat article here: http://bit.ly/24QqHHO