Structured query statement generation method and device, electronic equipment and medium

Abstract

The invention provides a structured query statement generation method and device, electronic equipment and a medium. The structured query statement generation method comprises the following steps: when a query request is received, extracting a statement from the query request, performing word segmentation processing on the statement, obtaining segmented words, generating a target list according tothe segmented words and a configured dictionary, wherein the dictionary represents a corresponding relationship between segmented words and segmented word types, the target list represents a corresponding relationship between target segmented words and target segmented word types; identifying the target segmented words in the target list as label values; generating a target syntax tree accordingto the label value and a pre-configured syntax tree framework, wherein the syntax tree framework represents a label, the target syntax tree represents a corresponding relationship between the label and a label value; obtaining a structured query statement template corresponding to the label based on the target syntax tree; and inputting the label value into the structured query statement templateto obtain a structured query statement corresponding to the statement.

Description

technical field [0001] The invention relates to the technical field of databases, in particular to a method, device, electronic equipment and medium for generating structured query statements. Background technique [0002] In order to facilitate data query, data is generally stored in the database in the form of a data table. For those who do not have the foundation of structured query language, it is impossible to organize structured query statements, and then the data results cannot be queried in the database. Quickly get data results, therefore, the method of converting natural sentences into SQL statements came into being. [0003] However, in the existing SQL statement conversion methods, they are mainly divided into seq2seq model methods and rule-based methods. In the method of the seq2seq model, a large number of training samples are required. However, such training samples are not only difficult to obtain, but also the acquisition efficiency is low, which leads to a...

AI Technical Summary

Problems solved by technology

In the method of the seq2seq model, a large number of training samples are required. However, such training samples are not only difficult to obtain, but also the acquisition efficiency is low, which leads to a lot of labor costs on the training samples. In addition, when the generated structured query statement is not accurate When , the seq2seq model cannot fix the corresponding errors, making the generation accuracy low; while in the rule-based method, it cannot cover all the user's sentence habits, so the user is required to input sentences according to certain rule templates, Bad for user experience, moreover, rule-based recall is low

Images