EAGER: Citizenly: Empowering Communities by Democratizing Urban Data Science

全国各地的政府一直在使用数据驱动的方法来推动城市运营和公民生活质量的改善。纽约、芝加哥和波士顿等大城市报告说，在利用数据解决城市问题方面取得了令人印象深刻的成果。像纽约州罗切斯特这样的小城市大多落在后面。数据在解决城市社区问题方面没有发挥很大作用。中型城市足够大，可以建立数据收集系统。但他们负担不起一个数据科学家团队来使用它们。由data.gov和许多其他组织提供的开放数据集包含原始数据。这些数据需要专门的技能来访问、理解和计算。大多数已建立的开放数据集最适合进行大数据分析。它们需要作出重大努力，将数据过滤到当地利益的具体情况。公民和社区领袖是开放数据的主要消费者。但他们无法以有意义的方式访问、破译和使用这些数据。虽然存在许多用于进行数据科学的工具和算法，但这些技术中的大多数不能简单地由任何人使用。数据科学民主化是指任何人，只要很少或没有技术专长，都可以从事数据科学。但必须向他们提供正确的数据和用户友好的工具。该项目将把城市数据科学交到公民和社区领导人手中。它将使公民参与城市政策的制定和地方问题的解决。社区领导人希望看到数据过滤到他们的社区层面。由此产生的分析应与公民相关。这项工作的一个主要影响是，它可以大大降低社区领导人和公民有效利用城市数据的门槛。特别重要的是社区青年的参与。他们将成为社区创新者，设计技术应用程序，以支持基于社区的自给自足战略。此外，该项目还将开发可供其他/类似中型城市复制的基础设施，并为其他领域的数据科学民主化铺平道路。该项目将开发创建Citizenly网络基础设施所需的基础科学和工程基础，以在公民和市政府中实现数据科学民主化。该项目是多学科的，它将汇集一个研究团队，在大数据分析，程序合成，主动学习和社会科学的专业知识。主要合作伙伴都在纽约州罗切斯特市，包括创新和战略举措办公室以及共同基础健康研究和分析办公室。该项目将解决以下关键的科学和技术挑战：（1）以社区为中心的数据基础设施：设计和实施轻量级公民传感器。为社区数据集设计一个高效的表达模型，并实现一个可扩展的抽取-转换-加载（ETL）系统来自动创建社区数据集。(2)以公民为中心的规划：为公民和社区领导者设计特定领域的语言，以访问社区数据集，并表达所需城市应用的意图和限制。(3)城市数据科学应用：使用Citizenly方法，开发：（a）空置土地对城市房地产价值影响的预测模型;（B）社会经济因素对健康影响的评估模型。Citizenly将整合社区数据和公共服务，沿着进行必要的系统和算法创新，为城市领导和市民提供下一代网络基础设施。Citizenly将通过为多元化的科学社区提供网络基础设施（CI）服务，广泛支持城市数据科学和基于社区的参与式研究。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。

Governments around the nation have been using data-driven approaches to drive improvements in city operations and citizens' quality of life. Large cities such as New York, Chicago, and Boston have reported impressive results from the use of data in addressing urban issues. Smaller cities such as Rochester, NY have mostly lagged behind. Data has not played a large role in addressing problems of urban neighborhoods. Midsize cities are large enough to have data collection systems in place. But they cannot afford a team of data scientists to make use of them. The open data sets, provided by data.gov and many other organizations, contain raw data. This data requires specialized skills to access, make sense of, and to compute with. Most established open data sets are most suitable for conducting large data analysis. They require significant efforts in filtering data down to specific scenario of local interests. Citizens and community leaders are key consumers of open data. But they are not able to access, decipher, and use this data in meaningful ways. While many tools and algorithms for conducting data science exist, most of these technologies cannot simply be used by anyone. Democratizing data science is the notion that anyone, with little to no technical expertise, can do data science. But they must be provided the right data and user-friendly tools. This project will put urban data science into the hands of citizens and community leaders. It will integrate citizens into the development of urban policy and solutions for local issues. Community leaders would like to see data filtered down to their community levels. The resulting analysis should be made relevant to citizens. A major impact of the work is that it can significantly lower the barrier to entry for community leaders and citizens to meaningfully leverage urban data. Of particular significance is the engagement of neighborhood youths. They will become neighborhood innovators, designing technology applications to support neighborhood-based self-sufficiency strategies. Moreover, the project will develop infrastructure that can be replicated for other/similar midsize cities as well pave the way to democratize data science in other domains.This project will develop the underlying scientific and engineering foundations necessary to create the Citizenly cyberinfrastructure to democratize data science amongst citizens and city governments. This project is multidisciplinary and it will bring together a team of investigators, with expertise in big data analytics, program synthesis, active learning, and social sciences. Key partners, all in the city of Rochester, NY, include the Office of Innovation and Strategic Initiatives and the Office of Research and Analytics at the Common Ground Health. The project will address the following key scientific and technical challenges: (1) Community-focused Data Infrastructure: Design and implementation of a lightweight citizen sensor. Design anexpressive and efficient model for community dataset and implement a scalable Extract-Transform-Load (ETL) system to automate its creation. (2) Citizen-centric Programing: Design domain specific language for citizens and community leaders to access community datasets and to express the intent and constraints for desired urban application. (3) Urban Data Science Applications: Using the Citizenly approach, develop: (a) predictive models for the influence of presence of vacant lots on city's property values; and (b) assessment models for health impact from socio-economic factors. Citizenly will integrate community data and common services along with necessary systems and algorithm innovations to provide next generation cyberinfrastructure for city leaders and citizens. Citizenly will enable research efforts broadly in urban data science and community-based participatory research by providing Cyberinfrastructure (CI) services to a diverse scientific community or communities.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.