Method and device for generating word feature vector, server, electronic equipment and storage medium

Abstract

The invention discloses a method for generating a word feature vector, and the method comprises the steps: obtaining a minimum training unit, and carrying out the word segmentation processing, thereby obtaining a target word set; respectively obtaining an initial feature vector of each target word in the target word set; training the target words in the target word set according to the difference between the sum of the initial feature vectors of all the target words in the target word set and the number of the target words in the target word set, and determining the word feature vector of each target word according to the training result. The invention further discloses a device and a server for generating the word feature vectors. According to the device and the method provided by the invention, the word feature vector can be quickly derived, so that the problems that a mainstream word vector generation method in the prior art is relatively low in training efficiency and prediction efficiency and needs to consume a large amount of memory and computing resources are solved.

Description

technical field [0001] The present invention relates to the technical field of natural language processing, and in particular, to a method for generating a word feature vector, an apparatus for generating a word feature vector, a server, an electronic device and a storage medium. Background technique [0002] Word feature vector (abbreviation: word vector) is the technology of choice for text vectorization in various natural language processing tasks, such as part-of-speech tagging, named entity recognition, text classification, document clustering, sentiment analysis, document generation, question answering systems, etc. A word vector can map words to a fixed dimension. Mathematically, a word vector can be expressed as a mapping relationship of f:x→y, where x is a word or paragraph text, and y is a fixed dimension vector after the mapping. [0003] The current mainstream word vector generation methods include Word2vec, GloVe, and BERT models. However, the Word2vec model ha...

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present invention provides a technical concept of training word feature vectors based on errors, and quickly derives word feature vectors through error calculation to solve the comparison between training efficiency and prediction efficiency of mainstream word vector generation methods in the prior art Problems that are low and consume a lot of memory and computing resources

Images