摘要
针对现有网构软件测试数据分类模型存在的分类相似度较低与召回率较低的问题,提出基于特征扩展的网构软件测试数据分类模型。对需分类的数据进行数据预处理,生成若干个网构软件测试数据的待测文件后,构建基于特征扩展的网构软件测试数据分类模型,使网构软件测试数据可以用基于特征扩展的向量表示,从而根据数据特征进行数据分类和存储。利用概念树计算数据相应权值和其表达能力,具体到每个网构软件测试数据进行概念描述。实验结果表明,基于特征扩展的网构软件测试数据分类模型能够有效提高分类数据的相似度和召回率。
Aiming at the problems of low classification similarity and low recall rate in the existing network architecture software test data classification models,a network software test data classification model based on feature extension is proposed.After preprocessing the data to be classified and generating several files to be tested,the network architecture software test data classification model based on feature extension is constructed,so that the network architecture software test data can be represented by the vector based on feature extension,so as to classify and store the data according to the data characteristics,and calculate the corresponding weight and its table of the data by using the concept tree The test data of each network architecture software is described in concept.The experimental results show that the classification model based on feature extension can effectively improve the similarity and recall rate of classification data.
作者
王钰
刘磊
WANG Yu;LIU Lei(Beijing Kedong Electric Power Control System Co.,Ltd.,Beijing 100085,China)
出处
《电子设计工程》
2021年第8期29-32,37,共5页
Electronic Design Engineering
基金
国家重点研发计划项目(2018YFB0904200)。
关键词
特征扩展
网构软件
测试数据
分类模型
feature extension
network software
test data
classification model
