Keyword extraction method and device and electronic equipment

Abstract

An embodiment of the invention provides a keyword extraction method and device and electronic equipment. The method includes: subjecting a to-be-processed text to word segmentation to obtain a plurality of word segments, and determining a word vector of each word segment; determining a label probability vector of each word segment according to the word vector of each word segment and a well trained BLSTM (bidirectional long short-term memory) network; aiming at each sentence of the to-be-processed text, subjecting each sentence to CRF decoding according to the label probability vector of eachword segment in each sentence to determine a classification label of each word segment in each sentence; determining word segments, with the classification labels being preset classification labels, in each sentence as keywords of the corresponding sentence; taking the keywords of each sentence in the to-be-processed text as keywords of the to-be-processed text. Network training is realized by construction of a neural network through the BLSTM network and CRF decoding, manual characteristic construction in a traditional method is avoided, and keyword extraction generalization capability is improved.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a keyword extraction method, device and electronic equipment. Background technique [0002] Keywords are words or phrases that can reflect the theme of the text, and are an important basis for people to quickly understand the content of the text and grasp the theme of the text. For example, in a news report, keywords can be used to grasp the theme and key content of the news report. In the thesis, the field and research topic of the thesis can be clarified through keywords. At present, keyword extraction technology has been widely used in information retrieval and text classification and other fields. The Internet has entered the era of Web 2.0. Many websites recommend objects of interest to users, such as videos, news, books, etc. They also need to use keyword extraction technology to satisfy users and deliver content in a more granular and scientific manner. , to achieve a...

AI Technical Summary

Problems solved by technology

[0004] However, the inventors found in the process of implementing the present invention that at least the following problems exist in the prior art: a large number of features need to be artificially constructed through a supervised method in order to train a network with better performance

However, artificially constructing features requires constructing different features for different fields, which has poor generalization ability and requires a lot of time and effort, and the quality of artificially constructed features directly affects the quality of the training network.

Images