Research on Recommended Method for Crowdsourcing Task of Mobile Application Based on Top-K Algorithm
Ying Liu, Tao Zhang, Kun Li, Hehe Gong · 2017
Crowdsourcing testers don't have to sign any contract, and the anonymous crowdsourcing testers determine the quality of tests.To match the recommended tasks with testers, a two phase task recommendation method based on Top-K algorithm was proposed.Category was introduced to reduce time complexity of Top-K algorithm.By classifying the tasks and calculating the category matching scores, the task types which were most suitable were obtained.After calculating the similarity between tester portrayal and tasks, the top K tasks were recommended from selected categories.Experiment shows that the proposed Top-K-Task algorithm can greatly improve the matching degree between testers and the recommended task. testin √ √To solve these problems, the paper studies the recommendation method of crowdsourcing testing tasks.First, an improvement on Top-K algorithm was proposed by introducing category.After calculating the matching score, the most suitable task types were selected.Then the top k tasks were