MRI: Acquisition of a Data Analytics Cluster for Computational Social Science

该奖项允许圣母大学的Jaaroslaw Nabrzyski博士和他的合作者通过收购计算社会科学数据分析集群（DACCSS）来实现社会科学的变革性研究。DACCSS提供的主要能力将加速现有的研究，这些研究因计算能力不足而受到限制，能够在以前无法达到的规模上进行分析和发现，孵化新的研究项目，并加强社会科学家与其同行之间的多学科合作。这种通过计算和先进的信息处理技术对广义社会现象的研究通常被称为计算社会科学（CSS）。在圣母大学（University of Notre Dame），多个系都有越来越多的教师利用CSS技术和能力。收购DACCSS为他们提供了超过2,600个CPU内核，5,300GB的RAM和60TB的高性能存储，这些存储在集群数据分析服务器上，连接着最先进的网络结构；运行世界上最先进的分析软件。社会科学数据集（如人口普查数据、详细调查、历史记录和电子设备/传感器日志）的规模、数量和可用性正在迅速增长。在高性能计算工具的推动下，随后的数据分析现在塑造了学者发现和与学生、同事和公众交流他们发现的方式。DACCSS系统将使CSS研究支持社会、行为和经济学（SBE）研究活动，如：社会学：分析复杂移动电话拓扑的社会网络模式和动态。心理学：多变量表型的全基因组分析，以研究精神和人格障碍的遗传基础。经济学：优化复杂和动态的多主体效用最大化模型，以理解家庭储蓄率的演变和影响，以及作为经济增长驱动因素的重要相关性。DACCSS对于这些和许多其他SBE研究努力中的变革性发现至关重要。社会科学院的出现要求在培养新社会科学学者方面进行创新；DACCSS的运营和培训管理计划将与CSS的教师、博士后研究人员和学生培训计划直接整合，并与同行大学共享。DACCSS也将在课堂上使用，并与调查人员讲授的许多课程相结合。daccss支持的研究将整合到多个现有的扩展项目中，以少数族裔和高中生为对象，如圣母大学暑期学者研究计算跟踪和NSF REU计算科学网站项目。

This award permits Dr. Jaaroslaw Nabrzyski and his collaborators at Notre Dame University to enable transformative research in the social sciences through the acquisition of a Data Analytics Cluster for Computational Social Sciences (DACCSS). The major capabilities provided by DACCSS will accelerate existing research throttled by insufficient computational capability, enable analysis and discovery at scales previously inaccessible, incubate new research projects and enhance multidisciplinary collaboration between social scientists and their peers. This investigation of broadly defined social phenomena through the medium of computing and advanced information processing technologies can be generally referred to as computational social science (CSS). At the University of Notre Dame, multiple departments have a growing number of faculty leveraging CSS techniques and capabilities. The DACCSS acquisition provides them with over 2,600 CPU cores, 5,300GB of RAM, and 60TB of high performance storage in clustered data analytics servers connected with a state of the art network fabric; running the world's most advanced analytics software. The size, number, and availability of social science datasets (such as census data, detailed surveys, historic records, and logs from electronic devices/sensors) is growing rapidly. Subsequent data analysis facilitated by high performance computational tools, now shapes the way that scholars discover and communicate their findings with students, colleagues, and the public. The DACCSS system will enable this CSS research supporting social, behavioral, and economics (SBE) research activities such as the following: Sociology: Analysis of social network patterns and dynamics with complex mobile phone topologies. Psychology: Genome-wide analyses of multivariate phenotypes to investigate the genetic underpinnings of mental and personality disorders. Economics: Optimization of complex and dynamic multi-agent utility maximizing models to understand the evolution and impact of household savings rates with important correlations as driving factors for economic growth. DACCSS is essential for transformative discovery in these and numerous additional SBE research endeavors.The emergence of CSS requires innovation in training for new social science scholars; the DACCSS operation and training management plan will have direct integration with a CSS training plan for faculty, post doctoral researchers and students of sufficient rigor to share with peer universities. DACCSS will also be accessible to the classroom with integration into numerous courses taught by the investigators. DACCSS-supported research will be integrated into multiple existing outreach programs to underrepresented minority and high school students such as the Notre Dame Summer Scholars Research Computing Track and the NSF REU Site program in Computational Science.