Dynamic Reward Scoring for Crowd-Computing Data Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing crowd-computing mechanisms face challenges in obtaining comprehensive and reliable member lists for large groups without a comprehensive member list, as participants are often incentivized to provide repetitive or incorrect answers due to simple payment structures.
Innovation Solution
A computer-implemented method and system that rewards participants with higher incentives for providing unique candidate members early, with validation thresholds and reward structures that prioritize early submissions and correct answers, ensuring that only validated members receive rewards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a simple fixed wage per input item is used, then the ease of operation is improved, but the reliability of the crowd-sourced data deteriorates
Solution Approach 1:
The patent changes the payment parameter from a fixed wage per item to a dynamic reward system based on uniqueness and timing. The reward amount is calculated as base_reward divided by the square root of the submission count, creating a parameter change that incentivizes early unique submissions while maintaining operational simplicity through automated calculation.
Solution Approach 2:
The system implements feedback by tracking submission counts and using this information to dynamically adjust rewards. Participants receive feedback on their reward amount based on how many times their input has been submitted, creating a self-regulating mechanism that improves data reliability without complex manual intervention.
2Productivity
If participants are rewarded for all submissions, then the productivity is improved, but the loss of information deteriorates due to repetitive answers
Solution Approach 1:
The patent applies parameter changes by modifying the reward calculation to account for submission frequency. The formula base_reward / sqrt(submission_count) creates a diminishing return effect that reduces incentives for repetitive submissions while still providing productivity incentives for unique contributions.
Solution Approach 2:
Instead of rewarding all submissions equally, the system inverts the approach by rewarding based on the inverse of submission frequency. This inversion ensures that the first unique submitters receive the highest rewards, naturally reducing redundancy while maintaining high productivity for novel contributions.
3Reliability
If early submissions are rewarded more highly, then the reliability is improved, but the loss of time deteriorates due to validation delays
Solution Approach 1:
The patent implements preliminary action by calculating and allocating rewards based on submission timing before validation is complete. The system determines reward amounts using the base_reward / sqrt(submission_count) formula immediately upon submission, preparing the reward structure in advance rather than waiting for validation to confirm accuracy.
Solution Approach 2:
The system uses self-service by automatically tracking submission counts and calculating rewards without manual intervention. The automated reward calculation based on submission frequency reduces the time loss associated with manual validation and reward distribution, allowing the system to self-regulate and incentivize reliable submissions efficiently.
Data Source
AI summary
Method, system, and computer program product are provided for scoring of crowd-computing inputs. A group of data is provided to crowd-computing participants and the participants are requested to provide candidate members or the group of data. The computer-implemented method performed includes: receiving an input by a participant, wherein the input is a candidate member; counting multiple inputs of the same candidate member by participants; validating a candidate member; rewarding the participants inputting the candidate member, with a higher reward for participants who input the candidate member earlier than other participants; and supplying the rewards to participants once the candidate member has been validated.