Automated Accuracy Assessment in Tasking System
Disclosed is a system (and process) for determining the accuracy of computerized tasks in a task batch. The system calculates a number of reviews to assess the accuracy of a task based on a source accuracy and a reviewer accuracy. The source accuracy is based factors calculated by a predictive model, the factors including a historical accuracy of an authoring user. The reviewer accuracy is based on a true positive rate and a true negative rate of one or more reviewers of the task batch. The system transmits sourced tasks to a same number of reviewers. The system collects reviews and assesses if the task passes review based on the collected number of reviews.
1 - 20 . (canceled)
21 . A method of determining an accuracy of computerized tasks, the method comprising:
receiving, by a computing system comprising one or more computing devices, an authored task from an authoring user, wherein the authored task is associated with analysis of one or more digital images;
determining, by the computing system, an author accuracy comprising a probability that the authored task has been correctly completed by the authoring user, wherein the author accuracy is based on a machine-learning prediction model configured to determine the author accuracy based on one or more factors comprising a historical accuracy of the authoring user;
responsive to the author accuracy not satisfying a quality threshold, transmitting, by the computing system, the authored task to one or more reviewers, wherein a number of the one or more reviewers receiving the authored task is based on the author accuracy;
responsive to transmitting the authored task to the one or more reviewers, receiving, by the computing system, one or more reviews of the authored task from the one or more reviewers, wherein each of the one or more reviews of the authored task comprises an indication of whether the authored task passes review; and
responsive to the author accuracy satisfying the quality threshold or a number of the one or more reviews satisfying a passed review threshold, generating, by the computing system, task data associated with the authored task.
22 . The method of claim 21 , wherein the transmitting, by the computing system, the authored task to the one or more reviewers comprises:
transmitting, by the computing system, the authored task to the one or more reviewers that have fewer tasks that have been completed and not reviewed than a threshold number of tasks.
23 . The method of claim 22 , wherein the threshold number of tasks is based on an aggregated historical task accuracy of the one or more reviewers.
24 . The method of claim 21 , further comprising:
determining, by the computing system, the author accuracy based on output from a machine-learning prediction model configured to use the one or more factors as input.
25 . The method of claim 21 , wherein the one or more factors comprises a behavior pattern of the authoring user, and wherein the behavior pattern of the authoring user comprises an amount of time for the authoring user to complete the authored task, a number of points drawn in the authored task, or a number of words submitted in the authored task.
26 . The method of claim 21 , wherein the number of the one or more reviewers is the same as the number of the one or more reviews.
27 . The method of claim 21 , wherein each of the one or more reviewers is associated with a respective status, and wherein the one or more reviews are weighted based in part on a status of the one or more reviewers.
28 . A non-transitory computer-readable medium storing instructions that when executed by a computer processor cause the computer processor to perform one or more operations comprising:
receiving an authored task from an authoring user, wherein the authored task is associated with analysis of one or more digital images;
determining an author accuracy comprising a probability that the authored task has been correctly completed by the authoring user, wherein the author accuracy is based on a machine-learning prediction model configured to determine the author accuracy based on one or more factors comprising a historical accuracy of the authoring user;
responsive to the author accuracy not satisfying a quality threshold, transmitting the authored task to one or more reviewers, wherein a number of the one or more reviewers receiving the authored task is based on the author accuracy;
responsive to transmitting the authored task to the one or more reviewers, receiving one or more reviews of the authored task from the one or more reviewers, wherein each of the one or more reviews of the authored task comprises an indication of whether the authored task passes review; and
responsive to the author accuracy satisfying the quality threshold or a number of the one or more reviews satisfying a passed review threshold, generating task data associated with the authored task.
29 . The non-transitory computer readable medium of claim 28 , wherein the transmitting the authored task to the one or more reviewers comprises:
transmitting the authored task to the one or more reviewers that have fewer tasks that have been completed and not reviewed than a threshold number of tasks.
30 . The non-transitory computer readable medium of claim 29 , wherein the threshold number of tasks is based on an aggregated historical task accuracy of the one or more reviewers.
31 . The non-transitory computer readable medium of claim 28 , further comprising:
determining the author accuracy based on output from a machine-learning prediction model configured to use the one or more factors as input.
32 . The non-transitory computer readable medium of claim 28 , wherein the one or more factors comprise a behavior pattern of the authoring user, the behavior pattern of the authoring user comprising at least one of an amount of time for the authoring user to complete the authored task, a number of points drawn in the authored task, and a number of words submitted in the authored task.
33 . The non-transitory computer readable medium of claim 28 , wherein each of the one or more reviewers is associated with a respective status, and wherein the one or more reviews are weighted based in part on a status of the one or more reviewers.
34 . A computing system comprising:
one or more processors; and
one or more tangible non-transitory computer-readable media that store instructions that when executed by the one or more processors cause the one or more processors to perform operations, the operations comprising:
receiving an authored task from an authoring user, wherein the authored task is associated with analysis of one or more digital images;
determining an author accuracy comprising a probability that the authored task has been correctly completed by the authoring user, wherein the author accuracy is based on a machine-learning prediction model configured to determine the author accuracy based on one or more factors comprising a historical accuracy of the authoring user;
responsive to the author accuracy not satisfying a quality threshold, transmitting the authored task to one or more reviewers, wherein a number of the one or more reviewers receiving the authored task is based on the author accuracy;
responsive to transmitting the authored task to the one or more reviewers, receiving one or more reviews of the authored task from the one or more reviewers, wherein each of the one or more reviews of the authored task comprises an indication of whether the authored task passes review; and
responsive to the author accuracy satisfying the quality threshold or a number of the one or more reviews satisfying a passed review threshold, generating task data associated with the authored task.
35 . The computing system of claim 34 , wherein the transmitting, by the computing system, the authored task to the one or more reviewers comprises:
transmitting, by the computing system, the authored task to the one or more reviewers that have fewer tasks that have been completed and not reviewed than a threshold number of tasks.
36 . The computing system of claim 35 , wherein the threshold number of tasks is based on an aggregated historical task accuracy of the one or more reviewers.
37 . The computing system of claim 34 , further comprising:
determining the author accuracy based on output from a machine-learning prediction model configured to use the one or more factors as input.
38 . The computing system of claim 34 , wherein the one or more factors comprise a behavior pattern of the authoring user, the behavior pattern of the authoring user comprises at least one of an amount of time for the authoring user to complete the authored task, a number of points drawn in the authored task, and a number of words submitted in the authored task.
39 . The computing system of claim 34 , wherein the number of the one or more reviewers is the same as the number of the one or more reviews.
40 . The computing system of claim 34 , wherein each of the one or more reviewers is associated with a respective status, and wherein the one or more reviews are weighted based in part on a status of the one or more reviewers.