查询语句处理方法、查询语句分类模型的构建方法、装置
By constructing a query statement classification model and using indicator features and semantic feature extraction modules to analyze SQL statements, the problem of low accuracy and efficiency in identifying abnormal SQL statements in existing technologies has been solved, achieving more efficient identification of abnormal SQL statements.
CN115438237BActive Publication Date: 2026-07-17JD DIGITS HAIYI INFORMATION TECHNOLOGY CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JD DIGITS HAIYI INFORMATION TECHNOLOGY CO LTD
- Filing Date
- 2022-08-31
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Current technologies rely on experience-based judgment to identify abnormal SQL statements, resulting in low efficiency and accuracy.
Method used
By constructing a query statement classification model, and using the indicator feature extraction module and the semantic feature extraction module, the query statements are analyzed from the aspects of monitoring indicators and semantic characteristics, respectively. The abnormal situations are determined by combining the first classification result and the second classification result.
Benefits of technology
It improves the accuracy and efficiency of identifying abnormal query statements and reduces reliance on experience-based judgment.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN115438237B_ABST
Abstract
本公开提供了一种查询语句处理方法，可以应用于计算机技术领域和人工智能技术领域。该查询语句处理方法包括：将待分类的查询语句信息的监控指标集合输入至指标特征提取模块，输出指标特征信息，其中，监控指标集合包括执行查询语句信息产生的监控指标；将指标特征信息输入至第一分类模块，输出查询语句信息的第一分类结果；将查询语句信息输入至第二分类模块，输出查询语句信息的第二分类结果；以及根据第一分类结果与第二分类结果，确定查询语句信息的异常情况。本公开还提供了查询语句分类模型的构建方法、装置、设备、介质及程序产品。
Need to check novelty before this filing date? Find Prior Art