Crowdsourcing task dynamic allocation method based on time window

Abstract

The invention provides a crowdsourcing task dynamic allocation method. The method comprises the steps of based on collected answers of tasks returned by workers in a previous time window, updating theposterior probability that all options of the tasks are correct answers through an answer decision method, updating the answer credibility of the tasks, and updating the quality of the workers; constructing a weighted bipartite graph, calculating the weight of each edge in the weighted bipartite graph by utilizing the task answer credibility continuous prediction method considering the field provided by the invention, and predicting the weight which is the answer credibility gain obtained by the task after a worker who predicts the edge connection answers the task of the edge connection; andcomparing the weights of all edges in the weighted bipartite graph, and preferentially allocating the task connected with the edge with the maximum weight to the corresponding worker. And the number of workers required by the task can be minimized under the condition of meeting the requirement of a task answer credibility threshold value, so that the cost is reduced.

Description

technical field [0001] The invention relates to the field of computer application technology, in particular to a method for dynamically allocating crowdsourcing tasks based on time windows. Background technique [0002] Crowdsourcing is to use the wisdom of the masses on the Internet to solve some problems that are difficult for computers to handle, such as judging whether two records belong to the same entity, language translation, image annotation, etc. Workers hope to get more remuneration, and task publishers hope to spend the least amount of money to ensure that the task is completed quickly and with high quality. Therefore, how to balance the relationship between the quality of task completion, required cost and completion time has always been a research field in the field of crowdsourcing. focus. In terms of task quality assurance, it is usually achieved by assigning a task to multiple workers to obtain redundant answers; in terms of cost and time control, either red...

AI Technical Summary

Problems solved by technology

[0006] (1) Fix the number of workers required for each task, without considering the difference in the difficulty of the task, resulting in a waste of manpower on simple tasks, and for difficult tasks, the quality of the completion may be low due to insufficient number of workers

[0007] (2) The impact of the order of arrival of workers on the number of workers required for the task is not considered, and the globally optimal allocation plan cannot be given

Images