CDI-Type II: Zooniverse - Conquering the Data Flood with a Transformative Partnership between Citizen Scientists and Machines

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。来自海洋学和生态学、天文学和经典研究等不同领域的科学家和研究人员面临着共同的挑战。随着计算机能力和技术的提高，我们可用的数据集的大小迅速增加。该项目的目标是开发一种新的方法，利用公民科学来释放现代大型数据集的知识发现潜力。例如，在之前的一个项目银河动物园中，公民科学家已经做出了重大贡献，他们的眼睛，模式识别技能和大脑可以解决需要人工输入的研究问题，并且已经成为计算过程的一部分。目前的银河动物园项目已经招募了20多万名参与者，他们提供了斯隆数字巡天的1亿多个星系分类。该项目建立在早期成功的基础上，开发了一种公民科学参与模式，不仅涉及简单的“点击”任务，而且还涉及参与者更先进的科学思维模式。作为该项目的一部分，将开发与机器学习工具和算法的共生关系，以便公民科学家的结果为改进算法提供丰富的训练集，从而为公民科学的参与模式提供信息。该项目的第一阶段将是从天体物理学、行星科学、动物学和经典研究中制定一系列试点项目。该项目的第二阶段将是制定一个框架-称为Zooniverse -以便利公民科学家。特别是，研究和机器学习社区将参与确定合适的项目和数据集，以整合到Zooniverse中。Zooniverse的最终目标是为大规模，基于互联网的公民科学创造一个可持续的未来，作为每个研究人员的一部分？的工具包，体现了计算思维的新范式，以创新和复杂的方式挖掘外行社区的精神资源，对我们产生新知识的能力产生深远的影响。该项目将使成千上万的公民参与真正的科学任务，从而使公众更好地了解科学，并通过学生的参与，引起对科学事业的兴趣。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). Scientists and researchers from fields as diverse as oceanography and ecology, astronomy and classical studies face a common challenge. As computer power and technology improve, the sizes of data sets available to us increase rapidly. The goal of this project is to develop a new methodology for using citizen science to unlock the knowledge discovery potential of modern, large data sets. For example, in a previous project Galaxy Zoo, citizen scientists have already made major contributions, lending their eyes, their pattern recognition skills and their brains to address research questions that need human input, and in so doing, have become part of the computing process. The current Galaxy Zoo project has recruited more than 200,000 participants who have provided more than 100 million classifications of galaxies from the Sloan Digital Sky Survey. This project builds upon early successes to develop a mode of citizen science participation which involves not only simple "clickwork" tasks, but also involves participants in more advanced modes of scientific thought. As part of the project, a symbiotic relationship with machine learning tools and algorithms will be developed, so that results from citizen scientists provide a rich training set for improving algorithms that in turn inform citizen science modes of participation. The first phase of the project will be to develop a portfolio of pilot projects from astrophysics, planetary science, zoology, and classical studies. The second phase of the project will be to develop a framework - called the Zooniverse - to facilitate citizen scientists. In particular, research and machine-learning communities will be engaged to identify suitable projects and data sets to integrate into Zooniverse.The ultimate goal with the Zooniverse is to create a sustainable future for large-scale, internet-based citizen science as part of every researcher?s toolkit, exemplifying a new paradigm in computational thinking, tapping the mental resources of a community of lay people in an innovative and complex manner that promises a profound impact on our ability to generate new knowledge. The project will engage thousands of citizens in authentic science tasks leading to a better public understanding of science and also, by the engagement of students, leading to interest in scientific careers.