PREDICTIVE MODEL OF TASK QUALITY FOR CROWD WORKER TASKS

摘要

Systems and methods of the present invention provide for server(s) assigning section or list item classifications to price list or business data extracted from a website. The server routes each new task verifying the classification to a crowd worker, and the server receives a completed. The server calculates a crowd worker score for each crowd worker based on each worker's quality scores according to the worker's review of the classifications on a worker user interface. The server generates a quality model for predicting a task quality score for the task, according to an error score for the crowd worker. If the error score in the quality model is below a predetermined threshold, the server transmits the completed task to a task reviewer's client for review.