Design Problems in Crowdsourcing: Improving The Quality of Crowd-Based Data Collection

Dissertation: Peter Organisciak

What's it about? Maximizing data quality in using paid crowds for objective and subjective encoding tasks. Read the Abstract.

Chapters

1. Introduction
2. Overview of Crowdsourcing
3. Design Facets of Crowdsourcing: A Typology for Information Science
4. Interpreting Objective Tasks for Paid Crowdsourcing
5. Designing Tasks for Objective Needs, part 2
6. Designing Tasks for Subjective Needs
7. Summary and Conclusions

Committee

• Michael B. Twidale, Professor, UIUC
• Miles Efron, Associate Professor, UIUC
• J. Stephen Downie, Associate Professor, UIUC
• Jaime Teevan, Principal Researcher, Microsoft Research

About this Repo

Dissertation

This project outlines Peter's dissertation research and writing.

Writing can be found in /documents, one document per chapter. It's written in Markdown (a text-only format), but is easily converted to laTex or HTML with Pandoc.

To Compile

	make pdf

Read the Makefile to understand how the markdown is converted, to Latex then to PDF.

Compiling Markdown files to laTeX

Pandoc is needed to convert Markdown to Latex -http://johnmacfarlane.net/pandoc/getting-started.html

cd document
pandoc -f markdown -t latex --smart --chapters -o 1_intro.tex 1_intro.md