A task content review timeliness guarantee method, device, equipment and medium

Abstract

The present application relates to a method, device, equipment and medium for ensuring the timeliness of task content review. The method includes: allocating different task ports for task content review according to the task type, and configuring different timeout alarm thresholds; during the review process, regularly refreshing and counting the number of tasks to be reviewed and the number of tasks that are about to time out for each business port, and constructing a timeout risk prediction model to perform multi-level timeout risk prediction on the tasks to be reviewed and send timeout alarms; periodically obtaining timeout alarm events and alarm response processing time, monitoring the task content review process in the entire link, and performing timeliness statistics during the monitoring process, adjusting and optimizing the review process based on the timeliness statistical results to further control and ensure the timeliness of task content review within the set period. The use of this method can effectively improve the efficiency of task review, speed up the task review time, reduce task timeouts and the waiting time for users to publish content, and enhance user experience.

Description

Technical Field

[0001] The present application relates to the technical field of network audit task processing, and in particular to a method, device, equipment and medium for ensuring the timeliness of task content audit. Background Art

[0002] With the development of Internet technology, the content on the Internet has gradually shifted from B-end (Internet products for corporate users) to C-end (consumer or individual user end). However, the content generated by C-end users (User Generated Content, referred to as UGC) has significant uncertainty, and the platform needs to review and filter the UGC content before publishing it. Under the general trend of C-end UGC content, it is particularly important for the platform to effectively review and filter UGC content, eliminate content that does not meet the requirements, quickly pass content without problems, and minimize the waiting time for user content to be published.

[0003] Under normal circumstances, user UGC content needs to go through the following process: after publishing → content flows to the platform's review system → platform review passes or rejects → content is displayed normally or deleted. During this period, if the content has been queuing in the platform's review system (there may be too much content to be reviewed, or the reviewer is temporarily away from work, or is reviewing other ports, such as nicknames, avatars, dynamics and other different series of ports, or other exceptional situations), due to the existence of various exceptional situations, the content posted by users cannot be reviewed in time, resulting in a poor user experience. Therefore, it is particularly important to avoid and handle such exceptions at the review system level.

[0004] To address this problem, the existing platform's task timeliness guarantee method often sets a task operation duration threshold to issue a task operation timeout warning, thereby improving the efficiency of task operation urging. However, this method does not take into account that different types of tasks may require opening separate pages for operation, and that the timeliness requirements and review standards for different types of tasks are also different. If the threshold is set too low, many complex tasks that are normally reviewed may be mistakenly considered to be timed out; on the contrary, if the threshold is set too high, simple tasks that are actually timed out may not be discovered in time, increasing the risk of false alarms and missed alarms of timeouts.

[0005] In addition, the platform has fluctuating task volumes. During specific times and periods, such as business promotion or hot events, the platform's incoming task volume surges and there is significant uncertainty. At the same time, auditors also face challenges in personnel management, such as temporary absences and intermittent breaks. With relatively fixed manpower, a surge in task volume or staff absence poses a huge challenge to the review time of content review. Existing methods do not monitor the entire life cycle of tasks, resulting in an intermittent increase in review time when there is a surge in task volume or when auditors leave their posts, causing a bad experience for platform users. Summary of the invention

[0006] Based on this, it is necessary to provide a task content review timeliness assurance method, device, equipment and medium to address the problems of high false alarm and missed alarm rates, lack of full-link task monitoring, and poor user experience on the platform in existing task timeliness assurance methods.

[0007] A method for ensuring timeliness of task content review, the method comprising:

[0008] Assign a batch of tasks stored in the set period to the current online auditor, who will review the task content by opening the business ports that match the various tasks. The business ports are configured with different timeout alarm thresholds based on the corresponding task types.

[0009] During the task content review process, the review page is refreshed regularly and the number of pending tasks of each business port and the number of tasks that are about to time out whose review time has reached the timeout alarm threshold of the corresponding task type are counted. The pending tasks of each business port are input into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task. When the timeout risk is predicted, the timeout alarm is sent to the current online auditor to request a response.

[0010] Periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and response processing time, and perform timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, adjust and optimize the current number of online auditors and the timeout risk prediction model, and further control and ensure the timeliness of task content review within the set period by adjusting and optimizing the results.

[0011] In one embodiment, the task types include: text review tasks, image review tasks, audio review tasks, and video review tasks.

[0012] In one embodiment, after the audit page is refreshed and the number of pending tasks of each service port and the number of tasks about to time out whose audit duration has reached the timeout alarm threshold of the corresponding task type are counted, the following is further included:

[0013] The number of tasks that are about to time out for each business port is displayed in red on the audit page, and the changes in the number are monitored. When the number of tasks that are about to time out does not improve within the set time, a desktop notification is sent through the browser to remind the current online auditor to pay attention to the audit time limit.

[0014] In one embodiment, the timeout risk prediction model predicts the multi-level timeout risk of the pending tasks of each service port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task, and sends a timeout alarm to the current online auditor to request a response when a timeout risk is predicted, including:

[0015] The overtime risk prediction model first calculates the hourly dimension review time of each task based on the historical hourly maximum review time and the historical hourly average review time of the current online auditor of each task, and performs a loose overtime risk prediction on the pending review tasks of each business port based on the hourly dimension review time of each task. If the prediction shows an overtime risk, a level 3 alarm is sent to the current online auditor; where hourly dimension review time = (historical hourly maximum review time + historical hourly average review time) / 2;

[0016] If the auditor responds to and handles the third-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared to before the update. If there is an improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further standard timeout risk prediction is performed; if the auditor does not respond to and handle the third-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and audit supervisor;

[0017] The timeout risk prediction model predicts the standard timeout risk of the current pending tasks of each business port based on the historical average hourly review time of the current online auditors of various tasks. If the prediction shows a timeout risk, a secondary alarm is sent to the current online auditor.

[0018] If the auditor responds to the second-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared with the previous update. If there is an improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further strict timeout risk prediction is performed; if the auditor does not respond to the second-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and audit supervisor;

[0019] The overtime risk prediction model conducts strict overtime risk prediction for the current pending tasks of each business port based on the historical fastest hourly review time of the current online auditors for each type of task. If an overtime risk is predicted, a level 1 alarm is sent to the current online auditor, and the audit supervisor is notified to conduct audit control until the number of pending tasks improves, the alarm level is lowered, and the alarm event is recorded.

[0020] In one embodiment, the prediction formula of the overtime risk prediction model for multi-level overtime risk prediction of pending tasks of each service port is expressed as:

[0021] ;

[0022] ;

[0023] in, Indicates the total review processing time of pending tasks for each business port. is the task processing time of each service port calculated according to the level of the timeout alarm. Indicates the current A pending task, The number of pending tasks for each business port. The maximum allowable review time for tasks at each service port. The task time of each service port has been stranded; if , predict that there is no risk of timeout for the pending tasks in the current batch; if , predicting that there is a risk of timeout for the pending review tasks in the current batch.

[0024] In one of the embodiments, the timeliness statistics include personal audit timeliness statistics and business port audit timeliness statistics.

[0025] A task content review timeliness guarantee device, the device comprising:

[0026] The task storage allocation module is used to allocate a batch of tasks stored in the set period to the current online auditor. The auditor will review the task content by opening the business ports corresponding to various tasks. Among them, the business ports are configured with different timeout alarm thresholds according to the corresponding task types.

[0027] The timeout risk prediction module is used to refresh and count the number of pending tasks of each business port and the number of tasks that are about to time out and whose review time has reached the timeout alarm threshold of the corresponding task type through the review page at regular intervals during the task content review process, and input the pending tasks of each business port into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task, and sends a timeout alarm to the current online auditor to request a response when the timeout risk is predicted;

[0028] The timeliness adjustment and optimization module is used to periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and response processing time, and perform timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, the current number of online auditors and the timeout risk prediction model are adjusted and optimized, and the timeliness of the task content review within the set period is further controlled and guaranteed by adjusting the optimization results.

[0029] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0030] Assign a batch of tasks stored in the set period to the current online auditor, who will review the task content by opening the business ports that match the various tasks. The business ports are configured with different timeout alarm thresholds based on the corresponding task types.

[0031] During the task content review process, the review page is refreshed regularly and the number of pending tasks of each business port and the number of tasks that are about to time out whose review time has reached the timeout alarm threshold of the corresponding task type are counted. The pending tasks of each business port are input into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task. When the timeout risk is predicted, the timeout alarm is sent to the current online auditor to request a response.

[0032] Periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and response processing time, and perform timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, adjust and optimize the current number of online auditors and the timeout risk prediction model, and further control and ensure the timeliness of task content review within the set period by adjusting and optimizing the results.

[0033] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0034] Assign a batch of tasks stored in the set period to the current online auditor, who will review the task content by opening the business ports that match the various tasks. The business ports are configured with different timeout alarm thresholds based on the corresponding task types.

[0035] During the task content review process, the review page is refreshed regularly and the number of pending tasks of each business port and the number of tasks that are about to time out whose review time has reached the timeout alarm threshold of the corresponding task type are counted. The pending tasks of each business port are input into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task. When the timeout risk is predicted, the timeout alarm is sent to the current online auditor to request a response.

[0036] Periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and response processing time, and perform timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, adjust and optimize the current number of online auditors and the timeout risk prediction model, and further control and ensure the timeliness of task content review within the set period by adjusting and optimizing the results.

[0037] Compared with the prior art, the above-mentioned task content review timeliness guarantee method, device, equipment and medium have the following technical effects:

[0038] 1. This application considers the differences in timeliness requirements and audit standards for different types of tasks, assigns different task ports for task content audit, and configures different timeout alarm thresholds, so that timeout warnings can be issued according to the specific audit requirements of each task, and warnings can be issued in advance before the task is about to time out, prompting auditors to speed up the progress and focus on urgent tasks, reduce the probability of task audit delays, and improve audit efficiency.

[0039] 2. This application constructs a timeout risk prediction model to perform multi-level timeout risk prediction on the pending tasks of each business port. It can predict the possible timeout risk of task content review in advance and send a timeout alarm to the current online auditor. The auditor can speed up the review progress and intervene in advance according to the alarm level, thereby avoiding the problem of timeout in processing efficiency caused by various unexpected situations.

[0040] 3. This application uses audit time statistics and full-link monitoring of the audit process to review the audit process based on the time statistics results, monitor the auditor's operating standards and processes, ensure the quality and efficiency of each auditor, and further ensure the timeliness of task audits.

[0041] This method can effectively improve the efficiency of task review, speed up the task review time, greatly reduce task timeouts, reduce the waiting time for users to publish content, and enhance the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A flowchart of a method for ensuring timeliness of task content review in an embodiment;

[0043] Figure 2 It is a structural block diagram of a task content review timeliness guarantee device in one embodiment;

[0044] Figure 3 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0046] In one embodiment, Figure 1 As shown, a method for ensuring the timeliness of task content review is provided, including the following steps:

[0047] Step S1, assign a batch of tasks stored in the set period to the current online auditor, and the auditor audits the task content by opening the business ports corresponding to various tasks; wherein the business ports are configured with different timeout alarm thresholds according to the corresponding task types.

[0048] Furthermore, the task types include: text review tasks, image review tasks, audio review tasks, and video review tasks.

[0049] It is understandable that due to the differences in timeliness requirements and audit standards for different types of tasks, it is necessary to configure corresponding alarm thresholds for different types of tasks and perform timeliness calculations, so that timeout warnings can be issued based on the specific audit requirements of each task, reducing the risks of misjudgment and missed judgments of timeouts.

[0050] Step S2, during the task content review process, the review page is refreshed and counted periodically to determine the number of pending tasks for each business port and the number of tasks that are about to time out and whose review time has reached the timeout alarm threshold of the corresponding task type. The pending tasks for each business port are input into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks for each business port based on the historical review time, average review time, and fastest review time of the current online auditor for each type of task. When a timeout risk is predicted, a timeout alarm is sent to the current online auditor to request a response.

[0051] Furthermore, after the audit page is refreshed and counted periodically for the number of tasks to be reviewed for each business port and the number of tasks that are about to time out whose review duration has reached the timeout alarm threshold of the corresponding task type, it also includes: displaying the number of tasks that are about to time out for each business port in red on the audit page, and monitoring changes in the number; when the number of tasks that are about to time out does not improve within the set time, a desktop notification is sent through the browser to remind the current online auditor to pay attention to the audit time limit.

[0052] It can be understood that by displaying the number of tasks that are about to time out in red, auditors can be reminded to pay attention to and focus on tasks that are about to time out, thereby reducing the probability of task review delays and improving review efficiency.

[0053] Furthermore, the specific process of the overtime risk prediction model for multi-level overtime risk prediction of pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each task includes:

[0054] The overtime risk prediction model first calculates the hourly dimension review time of each task based on the historical hourly maximum review time and the historical hourly average review time of the current online auditor of each task, and performs a loose overtime risk prediction on the pending review tasks of each business port based on the hourly dimension review time of each task. If the prediction shows an overtime risk, a level 3 alarm is sent to the current online auditor; where hourly dimension review time = (historical hourly maximum review time + historical hourly average review time) / 2;

[0055] If the auditor responds to and handles the third-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared to before the update. If there is an improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further standard timeout risk prediction is performed; if the auditor does not respond to and handle the third-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and audit supervisor;

[0056] The timeout risk prediction model predicts the standard timeout risk of the current pending tasks of each business port based on the historical average hourly review time of the current online auditors of various tasks. If the prediction shows a timeout risk, a secondary alarm is sent to the current online auditor.

[0057] If the auditor responds to the second-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared with the previous update. If there is an improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further strict timeout risk prediction is performed; if the auditor does not respond to the second-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and audit supervisor;

[0058] The overtime risk prediction model conducts strict overtime risk prediction for the current pending tasks of each business port based on the historical fastest hourly review time of the current online auditors for each type of task. If an overtime risk is predicted, a level 1 alarm is sent to the current online auditor, and the audit supervisor is notified to conduct audit control until the number of pending tasks improves, the alarm level is lowered, and the alarm event is recorded.

[0059] It can be understood that after the timeout risk prediction model issues a timeout alarm, the current online auditor needs to provide interactive feedback, and the feedback interaction is limited to a certain time range, such as 10s. If the auditor does not respond to the alarm event within 10 seconds, a first-level alarm will be directly triggered upward so that the audit supervisor can coordinate resources in a timely manner. The audit supervisor can also see the corresponding alarm triggering situation through the monitoring dashboard, and further assist in evaluating the current timeliness risk coefficient of the platform through the triggering ratio and triggering frequency based on the periodic alarm triggering ratio, alarm response processing time, triggering efficiency, etc. of the current platform. If the alarm triggering increases in a short period of time and the recovery time is long, it means that the auditor saturation is high in this time period and there is a possible audit timeout risk; on the contrary, if there is no alarm, there is no audit timeout risk.

[0060] Among them, the periodic alarm trigger ratio refers to: within a certain period of time, such as 7 days, the proportion of the platform's alarm trigger ports to the total audit ports. The alarm response processing time refers to the average time it takes for auditors and supervisors to complete the processing after the alarm occurs. This time is used to determine the upgrade time of the next level of alarm. For example, the slowest processing time for the third-level alarm is 2 minutes, but it has not been processed after 3 minutes, so a rapid second-level risk prediction is required. The trigger efficiency refers to whether the task review time is effectively accelerated when the alarm is triggered. Within a certain period of time, an AB comparison is made between alarms and no alarms, that is, A. No alarm is sent, B. Alarm is sent, and the timeliness of AB task processing is observed. If the processing time of the alarm sending end is significantly faster than that of the alarm not sending end, it is determined that sending the alarm is effective.

[0061] The timeliness risk coefficient means: if the alarm trigger ratio is high, or it is triggered frequently, or the processing time is long, and there are almost no alarms in the same time period, or it is much higher than the average value in the same time period, then the timeliness risk coefficient can be used to To evaluate, the specific algorithm is:

[0062] ;

[0063] in, F is the number of times the alarm is triggered within the current time (alarm trigger frequency), P It is the alarm trigger ratio, that is, the ratio of the platform's alarm trigger ports to the total audit ports. A is the average alarm response processing time; α , β , γ is the weight parameter of the corresponding factor, and its value is determined according to the number of tasks affected by historical alarms. RC The larger the value, the higher the risk of timeout in the current platform, and further processing is required. RC When the value is close to 0, it means that there is no timeout risk on the current platform.

[0064] Specifically, the prediction formula of the overtime risk prediction model for multi-level overtime risk prediction of pending tasks of each service port is expressed as:

[0065] ;

[0066] ;

[0067] in, Indicates the total review processing time of pending tasks for each business port. is the task processing time of each service port calculated according to the level of the timeout alarm. Indicates the current A pending task, The number of pending tasks for each business port. The maximum allowable review time for tasks at each service port. The task time of each service port has been stranded; if , predict that there is no risk of timeout for the pending tasks in the current batch; if , predicting that there is a risk of timeout for the pending review tasks in the current batch.

[0068] Among them, the task processing time corresponding to the third-level alarm level is: the average hourly processing time of the auditor for this business port within a certain period (such as 7 days), which is calculated by (highest value per hour + average value per hour) / 2; the task processing time corresponding to the second-level alarm level is: the historical average hourly review time of the auditor for this business port; the task processing time corresponding to the first-level alarm level is: the fastest hourly review time of the auditor for this business port.

[0069] Step S3, periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and the response processing time, and perform time statistics during the full-link monitoring process, adjust and optimize the current number of online auditors and the timeout risk prediction model based on the time statistics results, and further control and ensure the timeliness of the task content review within the set period by adjusting the optimization results.

[0070] Among them, full-link monitoring includes task dashboard timeliness monitoring, minute-level review volume monitoring, minute-level task addition monitoring, current online auditor number monitoring, current average review time monitoring, auditor entry into task port time monitoring, and task collection time monitoring. The audit supervisor can observe the status and timeliness of all pending tasks on the platform based on the monitoring dashboard, and can make reasonable plans based on the current inventory task volume and online auditors.

[0071] Time efficiency statistics are divided into personal audit time efficiency statistics and business port time efficiency statistics. Personal audit time efficiency statistics are used for refined management of auditors, and business port time efficiency statistics correspond to the audit time efficiency of various tasks. If the audit time efficiency of the entire business port is high and the alarm is frequent, it is necessary to consider whether the expected audit time is set unreasonably. If the personal audit time efficiency delay is high, it is necessary to review the auditor's operation nodes. Among them, the statistical cycles of time efficiency statistics include days, 10 minutes, half an hour, hours, etc. Through the 10-minute level, half an hour level, hour level, etc., it is possible to monitor the phased tasks well to prevent phased abnormalities or timeouts. Through the day level, the overall situation of each month or longer period can be seen. The main statistical indicators are: audit volume, average audit time, maximum audit time, minimum audit time, 0~1 minute completion rate, 0~3 minute completion rate, 0~5 minute completion rate, etc.

[0072] The audit supervisor can screen the abnormal data of personal audit time in the time statistics results, and review the audit operation steps of the auditor at that time based on the abnormal data of personal audit time. He can also find out the reasons for the timeout based on the review and adjust the number of auditors on duty, so as to further ensure the timeliness of the task audit.

[0073] Specifically, for example, the auditor starts recording when he opens the audit page, the business port currently opened by the auditor, records and counts the current number of tasks to be audited and the number of tasks that are about to time out on the business port, clicks to receive the task operation, receives the number of tasks, and if the image type requires the task resource to be enlarged, the audio and video task resources also need to monitor whether the auditor has completed the playback, and monitor the auditor's mouse track. Finally, the auditor clicks to approve or reject the audit task. The above process content will form a record of the auditor's entire process from task collection → task viewing → task operation. In order to review the entire process, the logic of the review is:

[0074] 1. Whether the number of tasks for review has increased sharply;

[0075] 2. The number of online reviewers at that time;

[0076] 3. Whether the auditor operates in accordance with the regulations;

[0077] 4. Time from when the task is put into the warehouse to when the task is collected, and the auditor’s operation status;

[0078] 5. Task received → Task operation time, auditor’s operation status.

[0079] Through the above review and response, we can further find out the reasons for the timeout, which will help further ensure the audit timeliness and optimize the process.

[0080] In summary, this application repeatedly confirms the review time target value through various means such as pre-timeout prediction, in-process evaluation / reminder / warning, and post-review / review. The review time can be roughly predicted when the task is put into the database. If it is not completed as expected during the process, a red reminder will be issued when it reaches the warning line, so that the auditor can handle it quickly. After the red reminder has not improved, a corresponding level of warning will be issued accordingly to urge the auditor, audit supervisor, etc. to further handle it; after the timeout situation is statistically analyzed and reviewed, the cause of the task timeout is found, and further optimized. The task content review time guarantee method provided by this application can effectively improve the efficiency of task review, speed up the task review time, greatly reduce task timeouts, reduce the waiting time for users to publish content, and enhance the user experience.

[0081] In one embodiment, Figure 2 As shown, a task content review timeliness guarantee device is provided, comprising:

[0082] The task warehousing allocation module 201 is used to allocate a batch of tasks that are stored in the set period to the current online auditor. The auditor reviews the task content by opening the business ports corresponding to each type of task; among them, the business ports are configured with different timeout alarm thresholds according to the corresponding task types.

[0083] The timeout risk prediction module 202 is used to periodically refresh and count the number of pending tasks for each business port and the number of tasks that are about to time out and whose review time has reached the timeout alarm threshold of the corresponding task type through the review page during the task content review process, and input the pending tasks for each business port into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks for each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task, and when the timeout risk is predicted, sends a timeout alarm to the current online auditor to request a response.

[0084] The timeliness adjustment and optimization module 203 is used to periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and the response processing time, and conduct timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, the current number of online auditors and the timeout risk prediction model are adjusted and optimized, and the timeliness of the task content review within the set period is further controlled by adjusting the optimization results.

[0085] For the specific definition of a task content review time limit guarantee device, please refer to the definition of a task content review time limit guarantee method above, which will not be repeated here. Each module in the above-mentioned task content review time limit guarantee device can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0086] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for ensuring the timeliness of task content review is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0087] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0088] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0089] Assign a batch of tasks stored in the set period to the current online auditor, who will review the task content by opening the business ports that match the various tasks. The business ports are configured with different timeout alarm thresholds based on the corresponding task types.

[0090] During the task content review process, the review page is refreshed regularly and the number of pending tasks of each business port and the number of tasks that are about to time out whose review time has reached the timeout alarm threshold of the corresponding task type are counted. The pending tasks of each business port are input into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task. When the timeout risk is predicted, the timeout alarm is sent to the current online auditor to request a response.

[0091] Periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and response processing time, and perform timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, adjust and optimize the current number of online auditors and the timeout risk prediction model, and further control and ensure the timeliness of task content review within the set period by adjusting and optimizing the results.

[0092] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0093] Assign a batch of tasks stored in the set period to the current online auditor, who will review the task content by opening the business ports that match the various tasks. The business ports are configured with different timeout alarm thresholds based on the corresponding task types.

[0094] During the task content review process, the review page is refreshed regularly and the number of pending tasks of each business port and the number of tasks that are about to time out whose review time has reached the timeout alarm threshold of the corresponding task type are counted. The pending tasks of each business port are input into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task. When the timeout risk is predicted, the timeout alarm is sent to the current online auditor to request a response.

[0095] Periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and response processing time, and perform timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, adjust and optimize the current number of online auditors and the timeout risk prediction model, and further control and ensure the timeliness of task content review within the set period by adjusting and optimizing the results.

[0096] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0097] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0098] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for ensuring the timeliness of task content review, characterized in that: The method comprises: Assign a batch of tasks stored in the set period to the current online auditor, who will review the task content by opening the business ports that match the various tasks; wherein the business ports are configured with different timeout alarm thresholds according to the corresponding task types; During the task content review process, the review page is refreshed regularly and the number of pending tasks of each business port and the number of tasks that are about to time out whose review time has reached the timeout alarm threshold of the corresponding task type are counted. The pending tasks of each business port are input into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task, and sends a timeout alarm to the current online auditor to request a response when a timeout risk is predicted; Periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and the response processing time, and perform time statistics during the full-link monitoring process. Based on the time statistics results, adjust and optimize the number of current online auditors and the timeout risk prediction model, and further control and ensure the timeliness of task content audit within the set period by adjusting and optimizing the results; The timeout risk prediction model predicts the multi-level timeout risk of the pending tasks of each service port according to the historical review time, average review time and fastest review time of the current online auditor of each task, and sends a timeout alarm to the current online auditor to request a response when a timeout risk is predicted, including: The overtime risk prediction model first calculates the hourly dimension review time of each type of task based on the historical hourly maximum review time and the historical hourly average review time of the current online auditor of each type of task, and performs a loose overtime risk prediction on the pending review tasks of each business port based on the hourly dimension review time of each type of task. If the prediction shows an overtime risk, a level 3 alarm is sent to the current online auditor; wherein, the hourly dimension review time = (historical hourly maximum review time + historical hourly average review time) / 2; If the auditor responds to and handles the third-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared with before the update. If there is improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further standard timeout risk prediction is performed; if the auditor does not respond to and handle the third-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and the audit supervisor; The timeout risk prediction model predicts the standard timeout risk of the current pending tasks of each business port according to the historical average hourly review time of the current online auditors of each type of task. If the prediction shows a timeout risk, a secondary alarm is sent to the current online auditor. If the auditor responds to and handles the second-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared with the previous update. If there is an improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further strict timeout risk prediction is performed; if the auditor does not respond to and handle the second-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and the audit supervisor; The overtime risk prediction model conducts strict overtime risk prediction for the current pending tasks of each business port based on the historical fastest hourly review time of the current online auditors for each type of task. If an overtime risk is predicted, a level 1 alarm is sent to the current online auditor, and the audit supervisor is notified to conduct audit control until the number of pending tasks improves, the alarm level is lowered, and the alarm event is recorded.

2. The method according to claim 1, characterized in that The task types include: text review tasks, image review tasks, audio review tasks and video review tasks.

3. The method according to claim 1, characterized in that: After the review page is refreshed and counted regularly, the number of pending review tasks for each business port and the number of tasks that are about to time out and whose review duration has reached the timeout alarm threshold of the corresponding task type are also included: The number of tasks that are about to time out for each business port is displayed in red on the audit page, and the change in number is monitored. When the number of tasks that are about to time out does not improve within the set time, a desktop notification is sent through the browser to remind the current online auditor to pay attention to the audit time limit.

4. The method according to claim 1, characterized in that The prediction formula of the overtime risk prediction model for multi-level overtime risk prediction of pending tasks of each service port is expressed as: ； ； in, Indicates the total review processing time of pending tasks for each business port. is the task processing time of each service port calculated according to the level of the timeout alarm. Indicates the current i A pending task, The number of pending tasks for each business port. The maximum allowable review time for tasks at each service port. The task time of each service port has been stranded; if , predict that there is no risk of timeout for the pending tasks in the current batch; if , predicting that there is a risk of timeout for the pending review tasks in the current batch.

5. The method according to claim 1, characterized in that The timeliness statistics include personal review timeliness statistics and business port review timeliness statistics.

6. A task content review timeliness guarantee device, characterized in that: The device comprises: The task storage allocation module is used to allocate a batch of tasks stored in the set period to the current online auditor. The auditor reviews the task content by opening the business ports corresponding to various tasks; wherein the business ports are configured with different timeout alarm thresholds according to the corresponding task types; The timeout risk prediction module is used to periodically refresh and count the number of pending tasks of each business port and the number of tasks that are about to time out and whose review time has reached the timeout alarm threshold of the corresponding task type through the review page during the task content review process, and input the pending tasks of each business port into the timeout risk prediction model. The timeout risk prediction model performs multi-level timeout risk prediction on the pending tasks of each business port according to the historical review time, average review time and fastest review time of the current online auditor of each type of task, and sends a timeout alarm to the current online auditor to request a response when the risk of timeout is predicted; A timeliness adjustment and optimization module is used to periodically obtain the timeout alarm events output by the timeout risk prediction model and the response processing time of the current online auditor, conduct full-link monitoring of the current online auditor's audit process based on the timeout alarm events and the response processing time, and conduct timeliness statistics during the full-link monitoring process. Based on the timeliness statistics results, the number of current online auditors and the timeout risk prediction model are adjusted and optimized, and the timeliness of the task content audit within the set period is further controlled by adjusting the optimization results; The timeout risk prediction model predicts the multi-level timeout risk of the pending tasks of each service port according to the historical review time, average review time and fastest review time of the current online auditor of each task, and sends a timeout alarm to the current online auditor to request a response when a timeout risk is predicted, including: The overtime risk prediction model first calculates the hourly dimension review time of each type of task based on the historical hourly maximum review time and the historical hourly average review time of the current online auditor of each type of task, and performs a loose overtime risk prediction on the pending review tasks of each business port based on the hourly dimension review time of each type of task. If the prediction shows an overtime risk, a level 3 alarm is sent to the current online auditor; wherein, the hourly dimension review time = (historical hourly maximum review time + historical hourly average review time) / 2; If the auditor responds to and handles the third-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared with before the update. If there is improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further standard timeout risk prediction is performed; if the auditor does not respond to and handle the third-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and the audit supervisor; The timeout risk prediction model predicts the standard timeout risk of the current pending tasks of each business port according to the historical average hourly review time of the current online auditors of each type of task. If the prediction shows a timeout risk, a secondary alarm is sent to the current online auditor. If the auditor responds to and handles the second-level alarm within the set time threshold, wait for the update of the number of pending tasks for each business port, and determine whether the subsequent number of pending tasks for each business port has improved compared with the previous update. If there is an improvement, the alarm is lifted and the alarm event is recorded; if there is no improvement, further strict timeout risk prediction is performed; if the auditor does not respond to and handle the second-level alarm within the set time threshold, a first-level alarm is directly sent to the current online auditor and the audit supervisor; The overtime risk prediction model conducts strict overtime risk prediction for the current pending tasks of each business port based on the historical fastest hourly review time of the current online auditors for each type of task. If an overtime risk is predicted, a level 1 alarm is sent to the current online auditor, and the audit supervisor is notified to conduct audit control until the number of pending tasks improves, the alarm level is lowered, and the alarm event is recorded.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images