A construction method of knowledge crowdsourcing platform based on incentive recommendation

Abstract

The invention discloses a method for constructing a knowledge crowdsourcing platform based on incentive recommendation. The method includes the following steps: constructing a labeling task completion mode under different incentive modes; constructing and training a recommendation model LRC for the labeling task completion mode based on different incentives; According to the user's Big Five personality test value, the model LRC is used to calculate the recommendation probability of different annotation task completion modes, and different annotation task completion modes are recommended according to the probability priority; according to the user's selected annotation task completion mode, assign and guide users to complete tasks, and obtain user tasks After completing the result, calculate the user reward in the corresponding task completion mode. By providing and recommending different task completion modes, the present invention satisfies the user's external and internal demands for task completion, reduces the money cost of the task issuer, and improves the crowdsourcing task completion quality.

Description

technical field [0001] The invention relates to computer technology, in particular to the field of artificial mixed intelligence, and in particular to a method for constructing a knowledge crowdsourcing platform based on incentive recommendation. Background technique [0002] Current deep learning models require massive amounts of data for training, resulting in a large number of labeling tasks. Different from expert annotation, it is a lower cost option to complete task annotation with the help of crowd wisdom. However, user groups with different personalities have different pursuits of autonomy, individuality and diversity, and have different motivations for task participation. The optimal task reward scheme should meet the incentive needs of different users. [0003] According to the analysis of existing crowdsourcing platforms, the current mainstream way of motivating users to participate and complete tasks is monetary incentives, which cannot meet the diverse incentive...

AI Technical Summary

Problems solved by technology

However, user groups with different personalities have different pursuits of autonomy, individuality, and diversity, and have different motivations for task participation. The optimal task reward scheme should meet the incentive needs of different users

[0003] According to the analysis of existing crowdsourcing platforms, the current mainstream way to motivate users to participate and complete tasks is monetary incentives, which cannot meet the diverse incentive needs of users with different personalities. Publishers need to bear high monetary costs

Images