WO2019200736A1

WO2019200736A1 - Operating method and device for crowdsourcing platform, computer device and storage medium

Info

Publication number: WO2019200736A1
Application number: PCT/CN2018/095318
Authority: WO
Inventors: 李佳琳; 刘鹏; 赵�怡; 王健宗; 肖京
Original assignee: 平安科技（深圳）有限公司
Priority date: 2018-04-17
Filing date: 2018-07-11
Publication date: 2019-10-24
Also published as: CN108734196A

Abstract

The present invention discloses an operating method and device for a crowdsourcing platform, a computer device and a storage medium. The method comprises: sending a test task to a crowdsourcing account; receiving feedback information of the crowdsourcing account and comparing the feedback information with a preset right answer to determine whether the answer of the crowdsourcing account is correct; if not, calling other test tasks in a test task library and sending to the crowdsourcing account until the answer of the crowdsourcing account is correct, and then sending a crowdsourcing task to the crowdsourcing account.

Description

Working methods, devices, computer equipment and storage media for crowdsourcing platforms

This application claims the priority of the Chinese Patent Application entitled "Working Methods, Devices, Computer Equipment and Storage Media for Crowdsourcing Platforms", filed on April 17, 2018, with the Chinese Patent Office, Application No. 2018103438779, the entire contents of which is hereby incorporated by reference. This is incorporated herein by reference.

Technical field

The present invention relates to the field of crowdsourcing, and more particularly to a method, apparatus, computer device and storage medium for a crowdsourcing platform.

Background technique

The crowdsourcing platform is a term coined in 2006 by Wired magazine to describe a new business model in which companies use the Internet to distribute work, discover ideas, or solve technical problems. When the enterprise verifies or improves image recognition, it will issue a photo proofreading task on the crowdsourcing platform in a crowdsourced manner to collect user feedback.

The user obtains the image proofreading task on the crowdsourcing platform, and the human identification system automatically gives the recognition result consistent with the image content. If it is inconsistent, the modification is performed, and then the “next” button is clicked to submit the answer. The crowdsourcing platform itself comes with the system to identify the results of the picture, so the user can choose to submit the system answer directly. Since the system verification principle is based on the same task assigned to at least three users, the three answers collected are automatically generated by cross-matching on the majority principle. When most users directly submit the wrong system answers directly, the system will This answer is judged to be correct (actually it may be wrong), so there is a user's brushing behavior loophole, which requires effective anti-brushing measures for real-time monitoring.

technical problem

SUMMARY OF THE INVENTION A primary object of the present invention is to provide a method, apparatus, computer device and storage medium for a crowdsourcing platform that can perform behavior testing on a crowdsourced account.

Technical solution

In order to achieve the above object, the present invention provides a method for operating a crowdsourcing platform, including:

Sending a test task to the crowdsourcing account according to a preset sending rule; wherein the answer displayed by the test task in the crowdsourcing account is a wrong answer, and the test task is in the same mode as a conventional crowdsourcing task;

Receiving feedback information of the crowdsourcing account for the test task, and comparing the feedback information with a preset correct answer, and determining whether the answer of the crowdsourcing account is correct;

If the answer to the crowdsourcing account is incorrect, the other test tasks are retrieved from the test task library and sent to the crowdsourcing account until the crowdsourcing account answers correctly, and then the crowdsourcing task is sent to the crowdsourcing account. Wherein, when the crowdsourcing account continuously answers the wrong number of times, the crowdsourcing task and the test task are suspended for the crowdsourcing account.

The invention also provides a working device for a crowdsourcing platform, comprising:

Sending a test unit, configured to send a test task to the crowdsourcing account according to a preset sending rule; wherein the test task displays an incorrect answer in the crowdsourcing account, and the test task is compared with a conventional crowdsourcing task The same pattern;

Receiving a comparison unit, configured to receive feedback information of the crowdsourcing account for the test task, and compare the feedback information with a preset correct answer to determine whether the answer of the crowdsourcing account is correct;

Executing an action unit, if the answer to the crowdsourcing account is incorrect, retrieving other test tasks in the test task library to continue to send to the crowdsourcing account until the crowdsourcing account answers correctly, and then sending the crowdsourcing task Giving the crowdsourcing account; wherein, when the crowdsourcing account continuously answers the wrong number of designated times, the crowdsourcing task and the test task are suspended for the crowdsourcing account.

The present invention also provides a computer device comprising a memory and a processor, the memory storing computer readable instructions, the processor executing the computer readable instructions to implement the steps of any of the methods described above.

The present invention also provides a computer non-transitory readable storage medium having stored thereon computer readable instructions that, when executed by a processor, implement the steps of any of the methods described above.

Beneficial effect

The working method, device, computer equipment and storage medium of the crowdsourcing platform of the present invention, because the test task and the conventional crowdsourcing task have the same mode, the operator of the crowdsourcing account does not know which of the tasks sent by the crowdsourcing platform, One is the test task, and the other is the crowdsourcing task. Therefore, the crowdsourcing platform can test whether the operator performs the brushing and so on without perceptually, and through this method, the user is continuously in the state of serious problem making, so that the crowdsourcing The platform maximizes the answers to valuable tasks. Moreover, if the crowdsourcing user continuously answers the wrong withdrawal task, it will notify the sending of the crowdsourcing task or the test task, etc., effectively preventing the crowding user's brushing behavior.

DRAWINGS

1 is a schematic flow chart of a method for operating a crowdsourcing platform according to an embodiment of the present invention;

2 is a schematic flow chart of a method for operating a crowdsourcing platform according to an embodiment of the present invention;

3 is a schematic flow chart of a method for operating a crowdsourcing platform according to an embodiment of the present invention;

4 is a schematic block diagram showing the structure of a working device of a crowdsourcing platform according to an embodiment of the present invention;

FIG. 5 is a schematic block diagram showing the structure of a working device of a crowdsourcing platform according to an embodiment of the present invention; FIG.

6 is a schematic block diagram showing the structure of a working device of a crowdsourcing platform according to an embodiment of the present invention;

FIG. 7 is a schematic block diagram showing the structure of an execution action unit according to an embodiment of the present invention; FIG.

FIG. 8 is a schematic block diagram showing the structure of a computer device according to an embodiment of the present invention.

BEST MODE FOR CARRYING OUT THE INVENTION

Referring to FIG. 1, an embodiment of the present invention provides a method for operating a crowdsourcing platform, including the steps of:

S1, sending a test task to the crowdsourcing account according to a preset sending rule; wherein, the answer displayed by the test task in the crowdsourcing account is a wrong answer, and the test task is in the same mode as a conventional crowdsourcing task;

S2. Receiving feedback information of the crowdsourcing account for the test task, and comparing the feedback information with a preset correct answer, and determining whether the answer of the crowdsourcing account is correct;

S3. If the answer to the crowdsourcing account is incorrect, the other test tasks are retrieved from the test task library and sent to the crowdsourcing account until the crowdsourcing account answers correctly, and then the crowdsourcing task is sent to the public. A package account; wherein, when the crowdsourcing account continuously answers the wrong number of times, the crowdsourcing task and the test task are suspended for the crowdsourced account.

As described in the above step S1, the above test task refers to a preset task stored in the test task library, the test task includes the task test question, the wrong answer and the correct answer, and the task test question and the wrong answer are sent when sent to the crowdsourcing account. Sent together, and the correct answer is used to compare with the feedback from the crowdsourced account. The above-mentioned crowdsourcing account is an account registered with the crowdsourcing platform, which can receive the tasks issued by the crowdsourcing platform, and process the crowdsourcing tasks, thereby obtaining rewards such as points, and can obtain corresponding points according to the number of points at the time of settlement. Money and other rewards. The above-mentioned conventional crowdsourcing task refers to a task such as an enterprise that issues a crowdsourcing task and actually needs crowdsourcing, such as sending a picture recognition task to a crowdsourcing account, to collect information on the accuracy of picture recognition, and the like. In this embodiment, the test task has the same mode as the crowdsourcing task, and the user corresponding to the crowdsourcing account can be tested without perception. For example, the mode of the crowdsourcing task is a picture and an explanatory text for the picture, then the test The task is also a picture and an explanatory text for the picture. The difference is that the interpretation of the picture in the crowdsourcing task may be correct, it may be wrong, and there is no preset correct answer, and the test task is correct. The explanatory text of the picture must be wrong and have the correct correct answer.

As described in the above step S2, the feedback information is the answer information of the user for the test task, and the answer information includes various situations. The first is that the user thinks that the answer displayed on the crowdsourcing account for the test task body is correct, and then Directly click on the feedback of the "next question"; the second is that the user finds that the answer displayed on the crowdsourced account for the test task body is wrong, and then modifies, and then submits the modified answer to the crowdsourcing platform, and then clicks "Next question" or feedback directly to the next question. In this embodiment, the preset correct answer refers to the correct answer set by the crowdsourcing platform for the task test. In the above process of “comparing the feedback information with the preset correct answer”, it generally involves techniques such as semantic analysis of the words. The meaning of the feedback information is the same as the meaning of the correct answer, and the crowdsourcing account is correctly answered. No need to be exactly the same word, otherwise, think that the crowdsourcing account answers incorrectly.

As described in the above step S3, the above comparison result is the result of the answer correctly or the answer error. A large number of different test tasks are stored in the above test task library. After sending a test task to the crowdsourcing account, the feedback answer of the crowdsourcing account is wrong, and then the other test tasks are transferred from the test task library to the crowdsourcing account until the crowdsourcing account answers correctly. The crowdsourcing task is then sent to the crowdsourced account. Through continuous testing, it can be judged whether the crowdsourcing account has the brushing behavior, or accidentally input the wrong answer. For example, if the user inputs the wrong answer continuously, it means that the crowdsourcing account has a brushing behavior; if the crowdsourcing account is only the first A test task is wrong, and the second test task is correct. It means that the user of the crowdsourcing account just made the wrong question, instead of deliberately brushing the score. In order to improve the answer accuracy of the crowdsourcing task, if the user has a brushing behavior, the sending to the crowdsourcing account crowdsourcing task is stopped. In other embodiments, the corresponding action is also performed according to the comparison result and the preset job rule. The preset job rule refers to a related rule set by the crowdsourcing platform, and is mainly for setting the comparison result, for example, If the answer is correct, continue to send the test task. If the test task is specified correctly for consecutively, the rules for the frequency of sending the crowdsourcing task are increased. The above-mentioned corresponding action is performed, that is, the crowdsourcing platform performs the operation according to the job rule. If the job rule is “An error is answered, the test task is continued to be sent”, then when the feedback information of the crowdsourcing account is different from the preset correct answer, the execution is continued. Send test task action.

In this embodiment, the method for calculating the answer of the crowdsourcing task by the crowdsourcing platform is a cross comparison method, which is as follows:

When the crowdsourcing platform distributes the same crowdsourcing task to n crowdsourcing accounts for verification, get the answers A1, A2, ... An, if there is the same answer A1= A2 = Ax, then B1 = 1 + 1 + 1 = 3, and so on, a total of k different answers, get B1, B2, ... Bk. If there is one and only one maximum B in B1, B2, ... Bk, and > 50%, the verification is successful, and the crowdsourcing platform adopts the majority answer corresponding to B. Conversely, if the verification fails, the system will re-deliver the crowdsourcing task and so on. The test task is directly compared with the preset correct answer, and no cross comparison is required. With the above cross-matching method, the crowdsourcing platform does not need to prepare the correct answer. It only needs to find the most answers with the same number of answers in multiple answers. The answer with the largest number of identical answers will be regarded as the correct answer. .

In this embodiment, the foregoing preset sending rule includes: sending a test task after sending a specified number of crowdsourcing tasks. For example, after sending 10 crowdsourcing tasks, send a test task.

In another embodiment, the foregoing preset sending rule includes: setting a corresponding number of test tasks according to the total number of crowdsourcing tasks, and transmitting the test tasks in a random distribution. For example, there are a total of 20 crowdsourcing tasks, and there are 4 test tasks that can be set up. These 4 test tasks are randomly interspersed in 20 crowdsourcing tasks for sending.

In still another embodiment, the foregoing preset sending rule includes: setting a corresponding number of test tasks according to the total number of crowdsourcing tasks, and transmitting the test tasks relatively densely at a beginning stage of sending the crowdsourcing task . For example, there are 100 crowdsourcing tasks, 20 test tasks are set up, 5 test tasks are interspersed in the top 10 crowdsourcing tasks, and the remaining 15 test tasks are randomly interspersed in the last 90 times. In the initial stage of sending the crowdsourcing task, the test task is sent relatively randomly, and the user corresponding to the crowdsourcing account can be found to be seriously responsible in time. For example, the crowdsourcing account answers the mistake when performing the test task. Obviously, the The user is not suitable to complete this type of crowdsourcing task, or there is a brushing behavior, and then stop sending the crowdsourcing task to the user.

In a specific embodiment, the foregoing preset transmission rules are as follows:

测试任务和众包任务总数N Total number of test tasks and crowdsourcing tasks N	测试任务分布 Test task distribution
N=[0-15] N=[0-15]	无测试任务 No test task
N=[16-40] N=[16-40]	随机分布5道测试任务（共5道测试任务） Randomly distributed 5 test tasks (5 test tasks in total)
N=[41-100] N=[41-100]	每15道分布一道测试任务（共4道测试任务） One test task per 15 channels (a total of 4 test tasks)
N=[101-200] N=[101-200]	每25道分布一道测试任务（共4道测试任务） One test task per 25 channels (a total of 4 test tasks)
N=[201-400] N=[201-400]	每20道分布一道测试任务（共10道测试任务） One test task per 20 channels (10 test tasks in total)
N>400 N>400	重复上述测试任务分布，值不发生变化 Repeat the above test task distribution, the value does not change

Referring to FIG. 2, in the embodiment, if the answer to the crowdsourcing account is incorrect, the other test tasks are retrieved from the test task library and sent to the crowdsourcing account until the crowdsourcing account answers correctly, and then Sending a crowdsourcing task to the crowdsourcing account; wherein, when the crowdsourcing account continuously answers the wrong specified number of times, after the step S3 of suspending the sending of the crowdsourcing task and the testing task to the crowdsourcing account, the method includes:

S311: Record the cumulative error number and/or the number of consecutive error corrections of the crowdsourcing account in response to the first specified length of time;

S312. Punve the crowdsourcing account according to the accumulated error number and/or the number of consecutive error answers.

As described in the above step S311, the first specified time length is generally several hours or one day, that is, the length of time from the start of the crowdsourcing task to the end of the settlement of each crowdsourcing account. Crowdsourcing tasks are generally settled on a daily basis. The above cumulative number of errors and answers refers to the number of times the crowdsourcing account has answered the wrong test tasks from the beginning of the crowdsourcing task until the settlement of the day. The above number of consecutive error-correcting numbers refers to the number of consecutive error-testing tasks using the crowdsourcing account. For example, if the user continuously answers the wrong five test tasks, the number of consecutive error-answering tasks is 5. In this embodiment, the cumulative number of errors and the number of consecutive errors can be recorded at the same time, or only one type of error can be recorded, which can be set according to requirements.

As described in step S312 above, a penalty rule is set, and the penalty rule is related to the cumulative number of errors and/or the number of consecutive answers.

In a specific embodiment, the error number calculation of the test task has two dimensions: a cumulative error number (K) and a continuous error number (Q), where K = {0, 1, 2, 3, 4, 5 },Q=={0,1,2,3,4}. The rule is: whenever the crowdsourcing account makes a wrong test task, the count of K is increased by 1, and the count of Q is incremented by one; whenever the crowdsourcing account is doing a test task, the count of K is unchanged, and the count of Q is reset. . Achieve the elimination of bad users in the shortest time and out of the crowdsourcing platform and correspondingly give different penalties. The specific counting and punishment process is as follows:

1. When the crowdsourcing account answers the wrong test task, mark K=1, Q=1 (the initial value of K and Q is 0);

2. Because when the crowdsourcing account fails in any one of the test tasks, the next question it will be will be the tester's task. At this time: (1) If the crowdsourcing account answers the second test task (ie, the error is not continuously answered) Test task), the count of K is unchanged, Q is reset to 0, that is, K=1, Q=0 at this time; (2) If the crowdsourcing account answers the wrong test task again (that is, two questions are wrong), and the error is marked at the same time. The number of questions K=K+1, Q=Q+1, that is, K=2 and Q=2 at this time;

3. According to the above rules, whenever the user completes a test task, the K and Q counts are updated in real time.

4. The different counts of K and Q correspond to different penalty coefficients DK and DQ, respectively. The specific correspondences are shown in the following two tables:

Correspondence between K and DK values

K K	0 0	1 1	2 2	3 3	4 4	5 5
DK DK	0 0	0 0	0 0	0.4 0.4	0.8 0.8	1 1

Correspondence between Q and DQ values

Q Q	0 0	1 1	2 2	3 3	4 4
DQ DQ	0 0	0 0	0.2 0.2	0.8 0.8	1 1

The score of DK and DQ means that the crowdsourcing platform will deduct the proportion of the task points on the day of the crowdsourcing account, and take the maximum value of the two as the current effective value of the penalty factor. For example, when DK=0.4, DQ=0.8, the system will take 0.8 as the effective value, deducting 80% of the user's task points for the day as a penalty. In addition, when DK=1, the system will clear all the task scores of the crowdsourcing account on the day of the crowdsourcing account, and re-count the number of tasks completed today; In addition to the singular and task points, the task assignment to the user of the brush is suspended within one hour.

Referring to FIG. 3, in the embodiment, after the step S311 of recording the cumulative error number and/or the number of consecutive error answers of the crowdsourcing account in the first specified time period, the method includes:

S313: Record a crowdsourcing account task operation situation within a second specified time length; wherein the second specified time length is greater than the first specified time length, and the second specified time length is the first Specifying a positive integer multiple of the length of time, the job situation including a cumulative number of error corrections and/or consecutive number of errors in the plurality of first time lengths of the crowdsourced account;

S314. Control a sending frequency sent to the crowdsourcing account according to the task status of the crowdsourcing account task.

As described in steps S313 and S314 above, it is to record the operation status of the above-mentioned crowdsourcing account, and then control the transmission frequency of the crowdsourcing task according to the operation situation. For example, if a person has multiple brushings within one month, the frequency of sending the task to the user is reduced. When the brushing condition is severe to a specified level (for example, if there is a brushing condition for 5 consecutive days), the The person's task account, etc. The brushing situation of the crowdsourcing platform can be further prevented.

In this embodiment, the step of retrieving another test task from the test task library to continue to be sent to the crowdsourcing account includes:

S3101: Search, according to the geographic area of the crowdsourcing account, a first test task library corresponding to the geographic area to which the belonging party belongs;

S3102: Retrieving a test task in the first test task library to continue to send to the crowdsourcing account;

S3103. Determine whether the test tasks in the first test task library are all used once;

S3104. If yes, the second test task library corresponding to the other geographical regions is replaced with each other.

Steps S3101, S3102, as described above S3103 and S3104, the test task library includes a plurality of test tasks, each test task library is set with a different test task, and each test task library is corresponding to a geographical area for distribution. For example, the test task in the A test task library is only for The crowdsourcing account in Guangzhou area is issued; when the test tasks in the A test task library are all used once, the test task library corresponding to other areas is replaced, such as the Guangzhou A test task library and the Hunan B test task library. , to reduce the multiple occurrences of the same test task in the same area, if the same question is answered twice by the same person, it will attract the attention of the user, and then know that it is a test task and take it seriously, and continue to brush points for other questions.

In this embodiment, the crowdsourcing platform also records the test task number sent to each crowdsourcing account to prevent the same test task from being sent to the same crowdsourced account multiple times. Specifically, each test task sets a number, records the number of each test task that each crowdsourcing account has answered, and if it needs to send a test task to the crowdsourcing account question, obtain the test that the crowdsourcing account has been tested. The task number, and then select the test task with its different number to send to the user for testing.

In this embodiment, the crowdsourcing platform continuously sends test tasks to all the crowdsourcing accounts participating in the answer at different times of each day, and then collects the answer questions, analyzes the answers of people in different regions, and then understands the different regions. The mental state of the person, thereby determining the time period for the crowdsourcing task in different regions. For example, users in the Guangzhou area have the best answer every morning at 9:00, and there is a serious brushing situation at 3 pm. In the subsequent sending task, the Guangzhou area raises the frequency of sending tasks at around 9:00 in the morning. Reduce the frequency of sending tasks, etc. at around 3 pm.

In this embodiment, it is also possible to perform long-term monitoring on each crowdsourcing account, and perform appropriate crowdsourcing tasks for the answering of crowdsourced accounts in different seasons in different regions. For example, the crowdsourcing account in Guangzhou exists in the summer. If the brushing behavior is serious, then in the summer, the frequency of crowdsourcing tasks for the crowdsourcing account in the region will be reduced, and in the winter, the answering efficiency will be higher, and in the winter, the frequency of crowdsourcing tasks in the region will be increased.

In this embodiment, the working method of the crowdsourcing platform further includes: interspersing a short text of the specified content between the crowdsourcing tasks. That is, in the process of sending crowdsourcing tasks, the crowdsourcing platform also intersperses and sends short essays between crowdsourcing tasks. The essays can be joke essays, soul chicken essays, etc. for users to alleviate fatigue. Further, in the short text interface such as joke or soul chicken soup, the corresponding next question or exit reading button will be set, and the user will receive a short message such as a joke essay or a soul chicken essay, and trigger the answer (click the next question or exit the reading button). The length of time, in turn, to determine the user's reading preferences, for example, the user will quickly click on the "exit reading" when the user encounters a cold joke essay, then after a period of analysis, no more jokes will be sent to the user.

The working method of the crowdsourcing platform of the embodiment of the present invention, because the test task and the regular crowdsourcing task have the same mode, the operator of the crowdsourcing account does not know which one of the tasks sent by the crowdsourcing platform is a test task, which One is the crowdsourcing task. Therefore, the crowdsourcing platform can test whether the operator performs the brushing and so on without perceptually, and through this method, the user can continue to be in the state of serious problem making, so that the crowdsourcing platform can obtain the maximum degree. Valuable task answers.

Referring to FIG. 4, an embodiment of the present invention further provides a working device of a crowdsourcing platform, including:

The sending test unit 10 is configured to send a test task to the crowdsourcing account according to a preset sending rule; wherein the test task displays an incorrect answer in the crowdsourcing account, the test task and a conventional crowdsourcing task The same pattern;

The receiving comparison unit 20 is configured to receive feedback information of the crowdsourcing account for the test task, and compare the feedback information with a preset correct answer to determine whether the answer of the crowdsourcing account is correct;

Executing the action unit 30, if the answer to the crowdsourcing account is incorrect, then retrieving other test tasks in the test task library to continue to send to the crowdsourcing account until the crowdsourcing account answers correctly, and then sending the crowdsourcing The task is to the crowdsourcing account; wherein, when the crowdsourcing account continuously answers the wrong number of times, the crowdsourcing task and the test task are suspended for the crowdsourcing account.

In the above sending test unit 10, the above test task refers to a preset task stored in the test task library, the test task includes a task test question, a wrong answer and a correct answer, and the task test questions and errors are sent to the crowdsourcing account. The answers are sent together, and the correct answer is used to compare the feedback with the crowdsourced account. The above-mentioned crowdsourcing account is an account registered with the crowdsourcing platform, which can receive the tasks issued by the crowdsourcing platform, and process the crowdsourcing tasks, thereby obtaining rewards such as points, and can obtain corresponding points according to the number of points at the time of settlement. Money and other rewards. The above-mentioned conventional crowdsourcing task refers to a task such as an enterprise that issues a crowdsourcing task and actually needs crowdsourcing, such as sending a picture recognition task to a crowdsourcing account, to collect information on the accuracy of picture recognition, and the like. In this embodiment, the test task has the same mode as the crowdsourcing task, and the user corresponding to the crowdsourcing account can be tested without perception. For example, the mode of the crowdsourcing task is a picture and an explanatory text for the picture, then the test The task is also a picture and an explanatory text for the picture. The difference is that the interpretation of the picture in the crowdsourcing task may be correct, it may be wrong, and there is no preset correct answer, and the test task is correct. The explanatory text of the picture may be wrong and has the correct correct answer.

In the above-mentioned receiving comparison unit 20, the feedback information is the answer information of the user for the test task, and the answer information includes various situations. The first type is that the user thinks that the answer displayed on the crowdsourcing account for the test task body is correct. And then directly click on the feedback of the "next question"; the second is that the user finds that the answer displayed on the crowdsourced account for the test task body is wrong, and then modifies, and then submits the modified answer to the crowdsourcing platform. Then click on "Next Question" or go directly to the feedback on the next question. In this embodiment, the preset correct answer refers to the correct answer set by the crowdsourcing platform for the task test. In the above process of “comparing the feedback information with the preset correct answer”, it generally involves techniques such as semantic analysis of the words. The meaning of the feedback information is the same as the meaning of the correct answer, and the crowdsourcing account is correctly answered. No need to be exactly the same word, otherwise, think that the crowdsourcing account answers incorrectly.

In the above-described execution action unit 30, the above-mentioned comparison result is the result of the answer correctly or the answer error. A large number of different test tasks are stored in the above test task library. After sending a test task to the crowdsourcing account, the feedback answer of the crowdsourcing account is wrong, and then the other test tasks are transferred from the test task library to the crowdsourcing account until the crowdsourcing account answers correctly. The crowdsourcing task is then sent to the crowdsourced account. Through continuous testing, it can be judged whether the crowdsourcing account has the brushing behavior, or accidentally input the wrong answer. For example, if the user inputs the wrong answer continuously, it means that the crowdsourcing account has a brushing behavior; if the crowdsourcing account is only the first A test task is wrong, and the second test task is correct. It means that the user of the crowdsourcing account just made the wrong question, instead of deliberately brushing the score. In order to improve the answer accuracy of the crowdsourcing task, if the user has a brushing behavior, the sending to the crowdsourcing account crowdsourcing task is stopped. In other embodiments, the corresponding action is also performed according to the comparison result and the preset job rule. The preset job rule refers to a related rule set by the crowdsourcing platform, and is mainly for setting the comparison result, for example, If the answer is correct, continue to send the test task. If the test task is specified correctly for consecutively, the rules for the frequency of sending the crowdsourcing task are increased. The above-mentioned corresponding action is performed, that is, the crowdsourcing platform performs the operation according to the job rule. If the job rule is “An error is answered, the test task is continued to be sent”, then when the feedback information of the crowdsourcing account is different from the preset correct answer, the execution is continued. Send test task action.

Referring to FIG. 5, in the embodiment, the working device of the above-mentioned crowdsourcing platform further includes:

a first recording unit 311, configured to record a cumulative error number and/or a continuous error number of the crowdsourcing account answer error within a first specified time period;

The punishment unit 312 is configured to punish the crowdsourcing account according to the accumulated error number and/or the number of consecutive error answers.

In the first recording unit 311 described above, the first specified time length is generally several hours or one day, that is, the length of time each crowdsourcing account ends from the start of the crowdsourcing task to the settlement. Crowdsourcing tasks are generally settled on a daily basis. The above cumulative number of errors and answers refers to the number of times the crowdsourcing account has answered the wrong test tasks from the beginning of the crowdsourcing task until the settlement of the day. The above number of consecutive error-correcting numbers refers to the number of consecutive error-testing tasks using the crowdsourcing account. For example, if the user continuously answers the wrong five test tasks, the number of consecutive error-answering tasks is 5. In this embodiment, the cumulative number of errors and the number of consecutive errors can be recorded at the same time, or only one type of error can be recorded, which can be set according to requirements.

In the above-mentioned penalty unit 312, a penalty rule is set, and the penalty rule is related to the cumulative number of errors and/or the number of consecutive answers.

Correspondence between K and DK values

K K	0 0	1 1	2 2	3 3	4 4	5 5
DK DK	0 0	0 0	0 0	0.4 0.4	0.8 0.8	1 1

Correspondence between Q and DQ values

Q Q	0 0	1 1	2 2	3 3	4 4
DQ DQ	0 0	0 0	0.2 0.2	0.8 0.8	1 1

Referring to FIG. 6, in the embodiment, the working device of the above-mentioned crowdsourcing platform further includes:

a second recording unit 313, configured to record a crowdsourcing account task operation situation within a second specified time length; wherein the second specified time length is greater than the first specified time length, and the second specified time The length is a positive integer multiple of the first specified length of time, and the job situation includes a cumulative number of error corrections and/or consecutive error corrections of the crowdsourced account over a plurality of the first length of time;

The sending control unit 314 is configured to control a sending frequency sent to the crowdsourcing account according to the crowdsourcing account task operation situation.

The second recording unit 313 and the transmission control unit 314 are units for recording the operation status of the above-mentioned crowdsourcing account, and control the transmission frequency of the crowdsourcing task according to the work situation. For example, if a person has multiple brushings within one month, the frequency of sending the task to the user is reduced. When the brushing condition is severe to a specified level (for example, if there is a brushing condition for 5 consecutive days), the The person's task account, etc. The brushing situation of the crowdsourcing platform can be further prevented.

Referring to FIG. 7, in the embodiment, the executing action unit 30 includes:

The finding sub-module 3101 is configured to search, according to the geographic area of the crowdsourcing account, a first test task library corresponding to the geographic area to which the belonging party belongs;

Retrieving a sub-module 3102, configured to retrieve a test task in the first test task library and continue to send to the crowdsourcing account;

The determining sub-module 3103 is configured to determine whether the test tasks in the first test task library are all used once;

The replacement sub-module 3104 is configured to replace the second test task library corresponding to the other geographical regions if the test tasks in the first test task library are all used once.

The above test task library includes a plurality of test tasks, and each test task library is set up with a different test task. For example, the test task in the A test task library is only for the crowdsourcing account in the Guangzhou area. Issuance; when the test tasks in the A test task library are all used once, the test task library corresponding to other areas is replaced, such as the Guangzhou A test task library and the Hunan B test task library are replaced, reducing the same test task. Multiple occurrences in the same area, if the same question is answered twice by the same person, it will attract the attention of the user, and then know that it is a test task and take it seriously, and continue to brush points for other questions.

In this embodiment, the job device of the crowdsourcing platform further includes a third recording unit for recording the test task number sent to each crowdsourcing account to prevent the same test task from being sent to the same crowdsourcing account multiple times. Specifically, each test task sets a number, records the number of each test task that each crowdsourcing account has answered, and if it needs to send a test task to the crowdsourcing account question, obtain the test that the crowdsourcing account has been tested. The task number, and then select the test task with its different number to send to the user for testing.

In this embodiment, the job device of the crowdsourcing platform further includes a sending and collecting unit, configured to send a test task to all the crowdsourced accounts participating in the question at different times of each day for consecutive days, and then collect the answer questions and analyze different The situation of people in the region, and then understand the psychological state of people in different regions, so as to determine the sending time of crowdsourcing tasks in different regions. For example, users in the Guangzhou area have the best answer every morning at 9:00, and there is a serious brushing situation at 3 pm. In the subsequent sending task, the Guangzhou area raises the frequency of sending tasks at around 9:00 in the morning. Reduce the frequency of sending tasks, etc. at around 3 pm.

In this embodiment, the job device of the crowdsourcing platform further includes a short text unit for interspersing the short text of the specified content between the crowdsourcing tasks. That is, in the process of sending the crowdsourcing task, the short message is interspersed between the crowdsourcing tasks, and the short text can be a joke short text, a soul chicken soup essay, etc. for the user to relieve fatigue. Further, the operation device of the crowdsourcing platform further includes a monitoring unit, which is configured to set a corresponding next question or exit reading button in a short text interface such as a joke or a chicken soup, and record the short text after the user receives the joke short text and the soul chicken soup essay. , the length of time that triggers the answer (click the next question or exit the button), and then judge the user's reading preferences. For example, if the user encounters a cold joke essay, he will quickly click “exit reading”, then analyze for a period of time. After that, no more jokes will be sent to the user.

Referring to FIG. 8, a computer device is also provided in the embodiment of the present invention. The computer device may be a server, and its internal structure may be as shown in FIG. 8. The computer device includes a processor, memory, network interface, and database connected by a system bus. Among them, the computer designed processor is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The non-volatile storage medium stores an operating system, computer readable instructions, and a database. The memory provides an environment for the operation of operating systems and computer readable instructions in a non-volatile storage medium. The database of the computer device is used to store data such as test tasks. The network interface of the computer device is used to communicate with an external terminal via a network connection. The computer readable instructions are executed by a processor to implement the processes of the various method embodiments described above.

In this embodiment, a computer non-volatile readable storage medium is further provided, where the computer readable instructions are stored, and when the computer readable instructions are executed by the processor, the processes of the foregoing method embodiments are implemented.

The above is only the preferred embodiment of the present invention, and is not intended to limit the scope of the invention, and the equivalent structure or equivalent process transformations made by the description of the invention and the drawings are directly or indirectly applied to other related The technical field is equally included in the scope of patent protection of the present invention.

Claims

A method for operating a crowdsourcing platform, comprising:

Sending a test task to the crowdsourcing account according to a preset sending rule; wherein the answer displayed by the test task in the crowdsourcing account is a wrong answer, and the test task is in the same mode as a conventional crowdsourcing task;

Receiving feedback information of the crowdsourcing account for the test task, and comparing the feedback information with a preset correct answer, and determining whether the answer of the crowdsourcing account is correct;

If the answer to the crowdsourcing account is incorrect, the other test tasks are retrieved from the test task library and sent to the crowdsourcing account until the crowdsourcing account answers correctly, and then the crowdsourcing task is sent to the crowdsourcing account. Wherein, when the crowdsourcing account continuously answers the wrong number of times, the crowdsourcing task and the test task are suspended for the crowdsourcing account.
The method for operating a crowdsourcing platform according to claim 1, wherein the preset sending rule comprises:

Send a test task after each specified number of crowdsourcing tasks are sent; or,

Setting a corresponding number of test tasks according to the total number of crowdsourcing tasks, and transmitting the test tasks randomly; or;

According to the total number of crowdsourcing tasks, a corresponding number of test tasks are set, and at the beginning of the sending crowdsourcing task, the test tasks are relatively randomly and randomly transmitted.
The method for operating a crowdsourcing platform according to claim 1, wherein if the answer to the crowdsourcing account is incorrect, another test task is retrieved from the test task library and sent to the crowdsourcing account. Until the crowdsourcing account answers correctly, then the crowdsourcing task is sent to the crowdsourcing account; wherein, when the crowdsourcing account continuously answers the wrong specified number of times, the step of suspending the sending of the crowdsourcing task and the testing task to the crowdsourcing account is suspended ,include:

Recording the cumulative number of incorrect answers and/or consecutive incorrect answers in the crowdsourced account within the first specified length of time;

The crowdsourced account is penalized based on the accumulated number of errors and/or consecutive number of answers.
The method for operating a crowdsourcing platform according to claim 3, wherein said step of recording the cumulative number of error corrections and/or consecutive number of error answers of said crowdsourced account answer error within said first specified length of time include:

Recording, in the second specified length of time, the crowdsourcing account task operation situation; wherein the second specified time length is greater than the first specified time length, and the second specified time length is the first specified time a positive integer multiple of the length, the job situation including a cumulative number of error corrections and/or consecutive number of errors in the plurality of first time lengths of the crowdsourced account;

The transmission frequency sent to the crowdsourcing account is controlled according to the crowdsourcing account task operation situation.
The method for operating a crowdsourcing platform according to claim 1, wherein the step of retrieving another test task from the test task library to continue to be sent to the crowdsourcing account comprises:

Searching, according to the geographic area of the crowdsourcing account, a first test task library corresponding to the geographic area to which the belonging is located;

Retrieving the test task in the first test task library continues to be sent to the crowdsourcing account.
The method for operating a crowdsourcing platform according to claim 5, wherein after the step of retrieving the test task in the first test task library to continue to be sent to the crowdsourcing account, the method comprises:

Determining whether the test tasks in the first test task library are all used once;

If yes, the second test task library corresponding to the other geographical regions is replaced with each other.
The method for operating a crowdsourcing platform according to claim 1, wherein the method further comprises:

Interspersed essays that send specified content between crowdsourcing tasks.
A work device for a crowdsourcing platform, comprising:

Sending a test unit, configured to send a test task to the crowdsourcing account according to a preset sending rule; wherein the test task displays an incorrect answer in the crowdsourcing account, and the test task is compared with a conventional crowdsourcing task The same pattern;

Receiving a comparison unit, configured to receive feedback information of the crowdsourcing account for the test task, and compare the feedback information with a preset correct answer to determine whether the answer of the crowdsourcing account is correct;

Executing an action unit, if the answer to the crowdsourcing account is incorrect, retrieving other test tasks in the test task library to continue to send to the crowdsourcing account until the crowdsourcing account answers correctly, and then sending the crowdsourcing task Giving the crowdsourcing account; wherein, when the crowdsourcing account continuously answers the wrong number of designated times, the crowdsourcing task and the test task are suspended for the crowdsourcing account.
The working device of the crowdsourcing platform according to claim 8, wherein the preset sending rule comprises:

Send a test task after each specified number of crowdsourcing tasks are sent; or,

Setting a corresponding number of test tasks according to the total number of crowdsourcing tasks, and transmitting the test tasks randomly; or;

According to the total number of crowdsourcing tasks, a corresponding number of test tasks are set, and at the beginning of the sending crowdsourcing task, the test tasks are relatively randomly and randomly transmitted.
The working device of the crowdsourcing platform according to claim 8, wherein the working device of the crowdsourcing platform further comprises:

a first recording unit, configured to record a cumulative number of error answers and/or consecutive error corrections of the crowdsourcing account answer error within a first specified time period;

And a penalty unit, configured to punish the crowdsourced account according to the accumulated error number and/or the number of consecutive answer errors.
The working device of the crowdsourcing platform according to claim 10, wherein the working device of the crowdsourcing platform further comprises:

a second recording unit, configured to record the crowdsourcing account task operation situation within a second specified time length; wherein the second specified time length is greater than the first specified time length, and the second specified time length Is a positive integer multiple of the first specified length of time, the job situation including a cumulative number of error corrections and/or consecutive error corrections of the crowdsourced account over a plurality of the first length of time;

And a sending control unit, configured to control a sending frequency sent to the crowdsourcing account according to the crowdsourcing account task operation situation.
The working device of the crowdsourcing platform according to claim 8, wherein the executing the action unit comprises:

a search sub-module, configured to search, according to the geographic area of the crowdsourcing account, a first test task library corresponding to the geographic area to which the belonging geographic area belongs;

Retrieving a sub-module for retrieving a test task in the first test task library to continue to send to the crowdsourcing account.
The working device of the crowdsourcing platform according to claim 12, wherein the executing the action unit further comprises:

a determining sub-module, configured to determine whether the test tasks in the first test task library are all used once;

The sub-module is configured to replace the second test task library corresponding to the other geographical regions if the test tasks in the first test task library are all used once.
The working device of the crowdsourcing platform according to claim 8, wherein the working device of the crowdsourcing platform further comprises:

A short text unit for interspersing essays that send specified content between crowdsourcing tasks.
A computer device comprising a memory and a processor, the memory storing computer readable instructions, wherein the processor implements a job method of a crowdsourcing platform when the computer readable instructions are executed, the crowdsourcing platform Working methods, including:

Sending a test task to the crowdsourcing account according to a preset sending rule; wherein the answer displayed by the test task in the crowdsourcing account is a wrong answer, and the test task is in the same mode as a conventional crowdsourcing task;

Receiving feedback information of the crowdsourcing account for the test task, and comparing the feedback information with a preset correct answer, and determining whether the answer of the crowdsourcing account is correct;

If the answer to the crowdsourcing account is incorrect, the other test tasks are retrieved from the test task library and sent to the crowdsourcing account until the crowdsourcing account answers correctly, and then the crowdsourcing task is sent to the crowdsourcing account. Wherein, when the crowdsourcing account continuously answers the wrong number of times, the crowdsourcing task and the test task are suspended for the crowdsourcing account.
The computer device according to claim 15, wherein the preset sending rule comprises:

Send a test task after each specified number of crowdsourcing tasks are sent; or,

Setting a corresponding number of test tasks according to the total number of crowdsourcing tasks, and transmitting the test tasks randomly; or;

According to the total number of crowdsourcing tasks, a corresponding number of test tasks are set, and at the beginning of the sending crowdsourcing task, the test tasks are relatively randomly and randomly transmitted.
The computer device according to claim 15, wherein if the answer to the crowdsourcing account is incorrect, another test task is retrieved from the test task library and sent to the crowdsourcing account until the public The package account is answered correctly, and then the crowdsourcing task is sent to the crowdsourcing account; wherein, when the crowdsourcing account continuously answers the wrong specified number of times, the steps of suspending the sending of the crowdsourcing task and the test task to the crowdsourced account include:

Recording the cumulative number of incorrect answers and/or consecutive incorrect answers in the crowdsourced account within the first specified length of time;

The crowdsourced account is penalized based on the accumulated number of errors and/or consecutive number of answers.
The computer apparatus according to claim 17, wherein said step of recording a cumulative number of error corrections and/or consecutive number of error corrections for said crowdsourced account answer error within a first specified length of time comprises:

Recording, in the second specified length of time, the crowdsourcing account task operation situation; wherein the second specified time length is greater than the first specified time length, and the second specified time length is the first specified time a positive integer multiple of the length, the job situation including a cumulative number of error corrections and/or consecutive number of errors in the plurality of first time lengths of the crowdsourced account;

The transmission frequency sent to the crowdsourcing account is controlled according to the crowdsourcing account task operation situation.
A computer non-volatile readable storage medium having stored thereon computer readable instructions, wherein the computer readable instructions are executed by a processor to implement a job method of a crowdsourcing platform, the operation of the crowdsourcing platform Methods, including:

Sending a test task to the crowdsourcing account according to a preset sending rule; wherein the answer displayed by the test task in the crowdsourcing account is a wrong answer, and the test task is in the same mode as a conventional crowdsourcing task;

Receiving feedback information of the crowdsourcing account for the test task, and comparing the feedback information with a preset correct answer, and determining whether the answer of the crowdsourcing account is correct;

If the answer to the crowdsourcing account is incorrect, the other test tasks are retrieved from the test task library and sent to the crowdsourcing account until the crowdsourcing account answers correctly, and then the crowdsourcing task is sent to the crowdsourcing account. Wherein, when the crowdsourcing account continuously answers the wrong number of times, the crowdsourcing task and the test task are suspended for the crowdsourcing account.
The computer non-volatile readable storage medium according to claim 19, wherein the preset sending rule comprises:

Send a test task after each specified number of crowdsourcing tasks are sent; or,

Setting a corresponding number of test tasks according to the total number of crowdsourcing tasks, and transmitting the test tasks randomly; or;

According to the total number of crowdsourcing tasks, a corresponding number of test tasks are set, and at the beginning of the sending crowdsourcing task, the test tasks are relatively randomly and randomly transmitted.