Keyword generation method and system based on variational inference theory

The invention provides a keyword generation method and system based on a variational inference theory. The method comprises the following steps of 1, preprocessing data; 2, constructing a model; step 3, model training; and 4, testing and evaluating the model. Aiming at the problem that the existing keyword generation method depends on a basic sequence-to-sequence framework to generate a target keyword and neglects the insufficient learning ability of copy and generation space representation, the invention analyzes a variational inference theory; variational inference is introduced into a keyword generation task, and a keyword generation model based on double hidden spaces is established, so that the purposes of generating high-quality keywords and self-adapting the number of pre-keywords according to text content by utilizing the characteristics of a Gaussian mixture module are achieved. 本发明提供了一种基于变分推断理论的关键词生成方法及系统，方法包括以下步骤：步骤1：数据预处理；步骤2：模型构建；步骤3：模型训练；步骤4：模型测试与评估。本发明针对现存的关键词生成方法依赖基本的序列到序列的框架来生成目标关键词，忽视了复制和生成空