Date of Award

Spring 1-1-2016

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Computer Science

First Advisor

Kenneth M. Anderson

Second Advisor

Leysia Palen

Third Advisor

Gloria Mark

Fourth Advisor

Tom Yeh

Fifth Advisor

Judith Stafford

Abstract

Crowdwork is a recent technique for mobilizing a large group of people to perform meaningful tasks for society on-line and to provide them with the tools they need to communicate, coordinate, and make progress. It is, however, an unexplored frontier in software development contexts in terms of having frameworks and tools that ease the development of crowdwork systems. This dissertation explores crowdwork systems in depth, surfacing the challenges that surround the creation of these systems via concerns raised by developers who have used and created crowdwork systems in the past. We use these concerns to identify a set of requirements for software frameworks that directly support the creation of crowdwork systems and then present a new framework—Crowdrouter—that meets these requirements. We evaluate the Crowdrouter framework by examining its versatility, i.e., with respect to its ability to support a wide variety of crowdwork- related workflows and tasks.

Share

COinS