Research On Generation Algorithm Of XPath Locator Based On Web Page Element Subject Recognition

In order to ensure the quality of web applications,web applications must be tested.As web application size and structural complexity increase,regression testing is often limited by resources such as time.Therefore,the efficiency of Web testing desired to be improved.More and more automated testing tools are being used extensively to increase the efficiency of web testing.However,in the regression test,the fixed web element locator is very fragile and often causes the test case to fail.To this end,this paper proposes an approach to crawling the content of the web page,using the topic model to identify the topic of the web element,and automatically generating an XPath locator.For the static page,this paper analyzes the reasons for the failure of the static page web element locator,expounds the topic vocabulary,the feature word extracted algorithm and the topic determination approach,introduces the XPath locator generated algorithm and realizes generating the XPath locator of the static page web element.In the static page,by limiting the topics and label names of the web elements.The approach has the advantages of less required samples and lower time complexity.For the dynamic page,this paper analyzes the reasons for the failure of the dynamic page web element locator,introduces the feature word extracted algorithm,the web element preprocess in the dynamic page,applies pLSA topic model and EM algorithm to determining the topic of web element,and introduces the XPath locator generated algorithm conjuncted with the web state-flow graph.The approach implements the web element XPath locator generation of the dynamic page.Finally,based on the research,this paper proposes three comprehensive problems and conducts related experiments.Experiments show that the XPath locator generation algorithm based on web element topic recognition can effectively identify web element topics,enhance the location ability of web elements,improve the robustness of web element locators,and improve the efficiency of testing to a certain extent.