자료유형 | 학위논문 |
---|---|
서명/저자사항 | Towards a Better Understanding of Peer-produced Structured Content Value. |
개인저자 | Hall, Andrew. |
단체저자명 | University of Minnesota. Computer Science. |
발행사항 | [S.l.]: University of Minnesota., 2019. |
발행사항 | Ann Arbor: ProQuest Dissertations & Theses, 2019. |
형태사항 | 112 p. |
기본자료 저록 | Dissertations Abstracts International 81-05B. Dissertation Abstract International |
ISBN | 9781088344699 |
학위논문주기 | Thesis (Ph.D.)--University of Minnesota, 2019. |
일반주기 |
Source: Dissertations Abstracts International, Volume: 81-05, Section: B.
Advisor: Terveen, Loren. |
이용제한사항 | This item must not be sold to any third party vendors. |
요약 | Over the last 30 years, peer production has created everything from software (e.g. Linux) to encyclopedia articles (e.g. Wikipedia) to geographic data (e.g. OpenStreetMap). In recent years, peer production has increased its focus on the production of structured (key-value pair) content. This content is designed to be consumed by applications and algorithms. This thesis explores two challenges towards generating content that is as valuable as possible to these applications/algorithms. The first challenge is unique to the context of peer-produced structured data and is focused on a tension between the core peer production ethos of contributor freedom and the need for highly-standardized data in order for applications/algorithms to effectively operate. To explore this tension between freedom and standardization, I qualitatively analyze the ways in which it surfaces and then quantitatively analyze its impact. For the second challenge, I compare how different levels of automation affect content value. Contributions in peer production come from manual editing, semi-automated tool editing, and fully-automated bot editing. I use two important lenses to study the value provided by these different types of contributions. Specifically, I study value by considering 1) the relationship between content quality and demand, and 2) problematic societal-level content biases (e.g. along male versus female, Global North versus Global South, and urban versus rural lines). While peer-production research has explored these two lenses of value in the past, it has not sought to develop a robust understanding in the context of structured content. To ensure that automated and manual contributions are effectively differentiated, I also develop a bot detection model. Finally, I provide implications based on my results. For example, my work motivates socio-technical tools that can reduce the manual effort required to contribute structured data and tools that direct effort towards in-demand content. |
일반주제명 | Computer science. |
언어 | 영어 |
바로가기 | ![]() |