Tasks Recommendation in Crowdsourcing based on Workers' Implicit Profiles and Performance History
Menna Maged Kamel, Alberto Gil‐Solla, Manuel Ramos-Carber · 2020
Crowdsourcing allows to build online platforms that make use of the power of human intelligence to complete tasks that are difficult to tackle for current algorithms. Current approaches to crowdsourcing adopt a methodology where tasks are published on specialized web platforms to a group of networked workers who can pick their preferred tasks freely on a first-come-first-served basis. Although this approach has several advantages, however it doesn't consider workers differences and capabilities. With the vast number of tasks posted by the requesters every day it's a challenging issue to satisfy both workers and requesters. In this paper, a crowdsourcing recommendation approach is proposed and evaluated that is based on a push methodology. This method aims to help workers to instantly find best matching tasks according to their interests and qualifications as well help the requesters to pick from the crowd the best workers for their desired tasks.