Using TensorFlow to deal with simple NLP problems

The current "AI" is the "big data" after another is about to be destroyed word, each company claims to exert artificial intelligence, as 4-5 years ago Big data, the industry called very loud, can not help but think of a foreigner said before:

Big Data is like teenage sex:everyone talks about it, nobody really knows how to does it, everyone thinks everyone else is Doing it, so everyone claims.

Now it seems that the above "big data" can be replaced with "AI", when we have not yet understood the big figures, AI began to lead the next trend. In the spirit of the follow-up attitude, I also try to spy on what exactly. Introduction

At present, both academia and industry, deep learning has been greatly sought after, especially in the Google open-source deep learning platform TensorFlow, but also for deep learning to fuel. Currently, TensorFlow is the most active in all open source projects on GitHub in the open source community, from rollout to the present, through several versions of the evolution that can be said to solve a large number of practical problems flexibly and efficiently. This paper mainly attempts to expound the simple application of tensorflow in natural language Processing (NLP), and let folks know tensorflow more emotionally.

Speaking of NLP, in fact, I am not very familiar with it, and have not had the relevant experience of NLP, this is my recent study of some of the accumulation of tensorflow, as a point. The internet is producing a lot of text and audio data every day, and by digging this data, we can do some more convenient applications, such as machine translation, speech recognition, POS tagging, and information retrieval, all of which belong to the NLP category. In the field of NLP, the language model is the most basic part, this article mainly revolves around the language model, first introduces its basic principle, then leads to the deep learning correlation model of Word vector (WORD2VEC), cyclic neural Network (RNN), short and long Time Memory network (LSTM), and introduces how to use TensorFlow implementation of the above model. Language Model

Language model is a probabilistic model, it is based on a corpus created, to get the probability of each sentence appear, popular point is to see a word is not normal people say, mathematically expressed as:

P (W) =p (W1W2...WT) =p (W1) p (W2|W1) p (w3|w1w2) ⋯p (wt|w1w2⋯wt−1)