基于句法分析的众包测试报告增强方法  

Crowdsourcing test report enhancement method based on syntactic parsing

作　　者：陈浩 黄松 史涯晴 骆润 Chen Hao;Huang Song;Shi Yaqing;Luo Run(Collegeof Command and Control Engineering,Army Engineering University of PLA,Nanjing,Jiangsu 210007,China)

机构地区：[1]陆军工程大学指挥控制工程学院,江苏南京210007

出　　处：《计算机时代》2021年第6期13-18,共6页Computer Era

基　　金：国家重点研发计划项目(2018YFB1403400);国家自然科学基金项目(61702544);中国博士后基金项目(2016M603031);江苏省自然科学基金项目(BK20160769,BK20141072)。

摘　　要：为了减少众包测试平台审查测试报告的数量,针对测试报告聚类后关联性高的报告进行句子层面的融合增强研究,对测试报告的描述语句构建依存句法树,然后通过分析句法树节点与关系边的一致性进行句子增强,最后输出结果语句。在实际众包测试报告数据集上的实验表明,该方法能够有效减少测试报告审查数量,并保证了增强报告的可读性。In order to reduce the number of manual review test reports in crowdsourcing test,this paper conducts a sentence-level fusion enhancement study for the reports with high relevance after clustering of test reports,builds a dependency syntax tree for the description sentences of the test report,and then bases on the consistency of the syntactic tree nodes and the relationship edges performs sentence enhancement,and finally output the result sentence.Experiments on the actual crowdsourcing test report data set show that this method can effectively reduce the number of test report reviews and ensure the readability of enhanced reports.

关 键 词：众包测试 测试报告 句法树 句子增强 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]