Re:How to Get into Natural Language Processing
by amelius on Saturday January 21, 2017 @05:58AM
(#13449639)
So what you are saying is that a computer would be able to understand sign-language more easily than spoken language?
Re:How to Get into Natural Language Processing
by visarga on Saturday January 21, 2017 @12:38PM
(#13451108)
Computers can transcribe spoken language into text at great accuracy, but they can't understand the meaning of text at the same level of accuracy yet. Meaning is much harder than simple transcription. Voice recognition is to speech like OCR to print. What we want is to speak to computers and have them understand what we mean, like humans. Such an AI would be able to carry a conversation, extract data from and reason over documents, or perform complex actions based on verbal commands. They would need to have a good physical and conceptual understanding of the world, otherwise they could not use reasoning.
In a way, vision is easier. Instead of discrete symbols (words) it's continuous signal which are much easier to interpret and generate from neural networks. By comparison, best language models are behind best image generation models (2-3 years behind, in my estimation).
For example, there are few applications of GANs to text, and many applications to images, GANs being the hottest thing in deep learning now. So you have to keep in mind that NLP is by and large still not solved. There is no decent conversational chat bot yet. We can reason over small pieces of text but that is far from full understanding. NLP at this level is hard.
What you can easily do now is to classify text, detect sentiment, entities, word vectors, grammatical parsing and summarization. All are low level stuff.