39 results about "Crowdsourced testing" patented technology

The invention discloses a method and device for processing crowdsourcing test data. The method comprises the following steps of: obtaining task description information of a to-be-tested application, wherein the task description information is used for indicating to carry out crowdsourcing test on the to-be-tested application; selecting a target tester set for testing the to-be-tested application from a to-be-selected tester set on the basis of a set of tester models and the task description information; and generating a crowdsourcing test strategy according to the target tester set and distributing the to-be-tested application to each tester in the target tester set according to the crowdsourcing test strategy. By adoption of the method and device, the efficiency of selecting testers in target tester sets can be improved, potential testers with good test resources and good test ability can be mined, and the potential testers can provide relatively good application test reports for application contractors so that the application contractors can carry out vulnerability patch on applications.

The invention discloses a web service crowdsourcing test task allocation method based on deep reinforcement learning, which comprises the steps of a deep reinforcement learning-based Web service testtask allocation model is trained according to data information in a worker pool and a task pool on a crowdsourcing platform; the crowdsourcing platform receives a test task submitted by a demander, and allocates the test task by using the trained Web service test task allocation model; and a worker receives and executes the task and feeds back a task test result to the crowdsourcing platform, andthe crowdsourcing platform transmits the test result fed back by the worker to the corresponding task demander. A deep reinforcement learning method DQN is used to train a Web service test task allocation model. The effect of distributing the Web service crowdsourcing test tasks in real time is achieved, it can be guaranteed to a certain extent that the test tasks can be handled by appropriate test personnel on the crowdsourcing test platform, and the test effect is improved.

The invention provides a crowdsourcing test task selection method and an electronic device, and the crowdsourcing test task selection method comprises the steps: collecting task demand description andcorresponding crowdsourcing reports of n open crowdsourcing test tasks, and calculating each descriptive term vector of each open crowdsourcing test task; extracting task selection features and inputting the task selection features into a task selection model through acquired activity information of a crowdsourcing test person and preference information and specialty information of each descriptive term vector of an i-th open crowdsourcing test task, and predicting the defect discovery probability of the crowdsourcing test person in each open crowdsourcing test task, i being greater than or equal to 1 and less than or equal to n; and selecting an optimal crowdsourcing test task based on the probability of discovering the defects. According to the crowdsourcing test task selection method,task information and crowd-testing personnel information on the crowd-testing platform are better represented; proper tasks are selected for crowd-testing personnel by mining and modeling historical records of the personnel; the task selection and exploration time of the personnel is shortened; the defect discovery efficiency of the personnel is improved; and the income of the personnel on the crowd-testing platform is improved.

The invention provides a crowdsourcing tester recommendation and crowdsourcing test method and an electronic device. The crowdsourcing tester recommendation and crowdsourcing test method comprises thesteps of: collecting crowdsourcing task demand description at a time point in a crowdsourcing software test process and historical crowdsourcing reports of to-be-recommended personnel; acquiring process context and resource context of each to-be-recommended personnel; and inputting the extracted features into a sorting learning model to obtain an initial recommended personnel sorting, and resorting the initial recommended personnel sorting based on specialty and equipment diversity contribution degrees to obtain a final recommended personnel sorting. Personnel recommendation is performed moreaccurately based on the current context information, and the recommended personnel can give consideration to both accuracy and diversity, so that the personnel can be dynamically planned in the crowddetection process, the defect detection rate is improved, the completion period of the crowd detection task is shortened, and the formation of a more efficient crowd detection service mode is promoted.

The invention discloses a test coverage calculation method and device based on an execution path tracking technology. The method comprises the following steps: collecting change data when a user triggers a control; sorting the change data according to the order of responding to the user triggering controls to obtain a sorted chart; counting execution paths based on the response chart, and performing gathering to generate a test execution path event set; and respectively calculating the test coverage rate and the trigger proportion of each execution path by adopting the test execution path event set. According to the method and device, test execution paths of different crowdsourcing workers can be tracked, and respective test coverage rates of the crowdsourcing workers can be counted and calculated and are fed back to the crowdsourcing workers in real time, so that the crowdsourcing workers can better understand the test defects of the crowdsourcing workers, and then more sufficient testing is carried out to ensure test coverage in crowdsourcing testing; and meanwhile, according to the triggering proportion of each event, a developer can better find the problem and the part in which people are interested, so that the mobile application is further perfected.

The invention provides a cooperation method for a crowdsourcing test platform, and the method has a reasonable guide mechanism and a real-time similar report recommendation mechanism, so crowdsourcingworkers carry out crowdsourcing software testing in a cooperation mode. The method is characterized in that the method comprises the steps: 1, analyzing the page structure of an application to be tested, enabling a defect report to carry a page label, and guiding the crowdsourcing workers to test by displaying page coverage of defects through Echarts Graph; 2, recommending similar defect reportsin real time through defect report label screening and text similarity calculation. 3, adding the user cooperation such as Fork modification, like clicking and treading of the defect reports, so thatthe defect reports are divided into a single type and a tree type, and crowdsourcing workers can audit the defect reports. The method has the beneficial effects that the test path is visualized through the page structure of the to-be-tested application, crowdsourcing workers are guided to reasonably test, and the test coverage rate is increased; through similar report recommendation, the number ofrepeated reports is reduced; and through a user cooperation mode, the quality of the defect report is iteratively improved.

The present invention provides a crowdsourcing testing method, comprising: the test client sends the characteristic information of the test client to the test server, and the test server selects the service to be tested according to the received characteristic information of at least one test client The client to be tested, the test server sends a list of services to be tested to the client to be tested, and the client to be tested tests the services to be tested in the list of services to be tested to obtain the service to be tested test results, and finally, the client to be tested sends the test results to the test server. This crowdsourcing testing method for testing network application services not only significantly improves the testing effect, but also effectively reduces the testing overhead.

The present invention discloses a crowdsourcing test amplification method based on test slices, which includes the following steps: 1) Using static information as a global guide for test amplification, running the test through dynamic symbol execution technology, and performing the operation information of the existing crowdsourcing test Analysis and statistics, in the form of test atoms, test slices, and test drives; 2) Identify and sort unexplored paths in the crowdsourcing test to be tested, and then select paths in order and solve them to obtain new test data; 3) Put the new test The data input is filled into the corresponding crowdsourcing test code template to obtain a new runnable test case, thus realizing the amplification of the crowdsourcing test results to be tested. The invention can effectively improve the test adequacy of crowdsourcing test codes, and at the same time generate test cases with high readability, adapt to crowdsourcing test scenarios, and fill the vacancy of crowdsourcing test amplification technology.

A method and device for processing crowdsourced test data, the method comprising: acquiring task description information of an application to be tested, the task description information being used to instruct crowdsourcing testing of the application to be tested; Set and the task description information, select the target tester set to test the application to be tested from the set of testers to be selected; generate a crowdsourced test strategy according to the set of target testers, and The application to be tested is delivered to each tester in the target tester set. By adopting this solution, the efficiency of selecting testers in the target tester set can be improved, and potential testers with good test resources and good testing ability can be discovered, and these potential testers can provide better application test reports for application developers. , so that the application developer can patch the application's vulnerabilities.

The invention provides a crowdsourcing tester recruitment method based on a social network. The crowdsourcing tester recruitment method comprises four stages of task segmentation, task release and propagation, propagation network construction and income distribution. The method comprises: firstly, dividing a public test task into small tasks which can be completed by a single tester; publishing the task to a crowdsourcing test platform, so that the task information is propagated in the social network; secondly, recording the superior of each tester by adopting an invitation code technology soas to form a propagation network; and finally, in a revenue distribution stage, performing revenue distribution according to the propagation network. According to the invention, the problem that enough testers cannot be recruited in a short time in the traditional crowdsourcing test is solved, and personnel guarantee is provided for crowdsourcing test. According to the method, the powerful personnel recruitment capability of the social network is utilized, enough workers are recruited for public survey in a short time, and meanwhile, a flexible income distribution mode is created, so that personnel recruitment has good motivation.