Word vector training method and device

Abstract

The invention provides a word vector training method and device. The method comprises the following steps: an internet webpage is captured, and training language materials are acquired and stored in a corpus; each training language material in the corpus is subjected to word segmentation, and an orderly word set corresponding to each training language material is obtained; a word list is established according to pre-collected user query logs; the training language materials stored in the corpus are distributed to nodes of a distributed word vector learning model; the distributed word vector learning model is configured to perform periodic word vector training on each word in the word list, and the word vector corresponding to the word in the word list is obtained. According to the word vector training method and device, the word vectors obtained through training can be well applied to search business, and fast iterative high-quality word vector training can be realized.

Description

technical field [0001] The present invention relates to the technical field of the Internet, in particular to a word vector training method and device. Background technique [0002] In Internet applications, a very important issue is how to convert natural language into a data expression form that computers can understand. The most important step to solve this problem is to find a way to digitize natural language symbols. At present, the deep learning (Deep Learning, DL) method is commonly used. In DL, the "Distributed representation" distributed representation method is used, and each word is represented as a low-dimensional real number vector, which is the word vector corresponding to the word. Word vectors are born from this. It is understandable that word vectors are vectors used to express words in natural language, so that they are suitable for Internet applications. For example, word vectors can be used in many natural language learning processing (Natural Learning ...

AI Technical Summary

Problems solved by technology

This word vector training method uses a stand-alone mode, and its training rate is low, especially it is difficult to apply to business scenarios with a very large amount of data

In addition, this word vector training method is a universal training method, which does not consider the particularity of specific business scenarios, so the training effect is not good in specific business scenarios

Images