The Stanford Data Science Collaboratory

数据驱动的调查是科学和发现的各个方面的关键，基于数据的决策正在成为社会的一部分。当目标是获得相关的、有效的、可复制的科学见解时，挑战和应对挑战的重要性尤为关键。斯坦福大学数据科学合作实验室将通过创建一个由教师、博士后学者、学生和研究人员组成的社区来应对这些挑战，这些社区将利用数据科学方法和领域知识来解决紧迫的问题。 在合作实验室，数据科学家将与其他领域的学者密切合作，这些学者依赖于大型，准确，可靠的数据集和数据科学技术。 合作实验室将促进研究人员的工作，他们研究与数据收集和使用有关的伦理问题，并将使用数据来解决社会和科学问题。一个标志将是对数据进行彻底的验证，并对证据进行仔细的统计校准，以避免可能产生不利后果的误解。 合作实验室的第二个主要目标是培养公民对数据科学的了解：大学有义务确保下一代了解如何解释和学习数据，以及如何收集和管理数据。合作实验室确定了一组五个备受瞩目的，高影响力的项目，领域科学家认为这些项目很重要，他们认为如果不改变处理数据集的方式，就无法取得进展。前两个问题涉及人类与环境之间的可持续关系，即（1）在气候变化中管理珊瑚礁的问题和（2）减少金枪鱼供应链中的非法捕捞和强迫劳动。为了取得进展，该项目将利用新的数据来源：卫星遥感、基于声景的地面监测站以及跟踪生物多样性和进化的基因组测量。 其他三个项目是关于社会的断裂和走向可持续发展的步骤：（3）如何理解美国贫困的决定因素; (4)如何检测和跟踪数字媒体中的政治框架;（5）如何开发支持个人之间公平对待的数据科学工具。 公共数据流（例如，社交媒体应用程序、维基百科和维基数据以及中等分辨率的卫星图像）以及私营部门的数据（例如，手机记录、脸书活动、互联网搜索查询、无人机图像和高分辨率卫星数据）将有助于了解造成贫困的机制。为了实现研究目标，合作实验室将激励教师，学生和博士后走到一起，通过支持合作研究团队和头脑风暴工作组来寻找新的数据科学解决方案。合作实验室还将通过提供实践指导的科学研究经验来吸引本科生。为了扩大斯坦福大学以外的合作，合作实验室将接待外部访问者，并邀请科学家到校园参加年度研讨会。该项目是美国国家科学基金会利用数据革命（HDR）大创意活动的一部分。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Data-driven inquiry is key to all aspects of science and discovery, and data-based decisions are becoming integral to society. The challenges and the importance of meeting them are especially critical when the goal is to obtain relevant, valid, reproducible scientific insights. The Stanford Data Science Collaboratory will confront these challenges by creating a community of faculty, postdoctoral scholars, students, and research fellows that leverage data science methods and domain knowledge to tackle pressing problems. In the Collaboratory, data scientists will work closely with scholars from other fields who rely on large, accurate, dependable datasets and data science techniques. The Collaboratory will foster the work of researchers who study the ethical issues related to data collection and use, and will use data to solve societal and scientific problems. A hallmark will be thorough validation of data and a careful statistical calibration of the evidence to avoid misinterpretations that could have adverse consequences. A second major goal of the Collaboratory is the growth of a citizenry literate in data science: universities have an obligation to ensure the next generation understands how to interpret and learn from data, and how to collect and manage it.The Collaboratory identifies a set of five high-profile, high-impact projects that domain scientists deem important, and where they believe they are unable to make progress without a paradigm shift in the way they approach data sets. The first two concern a sustainable relation between humans and the environment, namely, (1) the problem of managing coral reefs in a changing climate and (2) reducing illegal fishing and forced labor in tuna supply chains. To make progress, the project will leverage new data sources: satellite remote sensing, ground monitoring stations based on soundscape, and genomic measurements that track biodiversity and evolution. The other three projects are about fractures in society and steps towards a sustainable one: (3) How to understand the determinants of poverty in the U.S.; (4) How to detect and track political framing in digital media; and (5) How to develop data science tools that support equitable treatment between individuals. Public data streams (e.g., social media apps, Wikipedia and Wikidata and moderate-resolution satellite imagery) as well as private-sector data (e.g., cell phone records, Facebook activity, internet search queries, drone imagery and fine-resolution satellite data) will inform understanding of the mechanisms causing poverty. To meet the research goals, the Collaboratory will incentivize faculty, students and postdocs to come together to find new data science solutions by supporting collaborative research teams and brainstorming working groups. The Collaboratory will also engage the undergraduate community by providing hands-on guided scientific research experience. To enlarge collaboration beyond Stanford, the Collaboratory will host outside visitors and invite scientists to campus for an annual symposium.This project is part of the National Science Foundation's Harnessing the Data Revolution (HDR) Big Idea activity.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.