NCS-FO: Cognitive maps as a framework for organizing relationships in large-scale real social networks

人类的生活在我们的社交网络的广阔版图中上演。成功的社交导航可以帮助缓冲孤独和负面互动，但它需要了解谁与谁相连：潜在的社会关系结构(例如，集团或中心)。该项目由布朗大学的一组研究人员领导，目的是揭示人们如何学习和推理现实世界中的社交网络的行为和神经特征。尽管了解人们如何学习关系型社会知识很重要，但关于大脑如何组织这种知识仍有许多未知之处。空间认知研究为这个问题提供了一个窗口：一长串动物研究表明，大脑在一张“认知地图”中代表着有关物理空间的信息，这张地图绑定了关于世界上实体及其关系的信息。认知地图已被证明可以在新的或不断变化的环境中加快学习速度。该项目测试的是，在精神上浏览社交网络时，人们是否会招募到类似于外部物理空间导航的认知地图表征。通过将先进的人体成像方法、计算建模、社交网络科学和纵向抽样相结合，研究人员将研究一大批一年级本科生在一学年期间发展新友谊的过程，以调查关于现实世界社交网络的新兴知识的行为和神经特征。研究人员将在这个复杂而动态的环境中识别社会信息的神经表征。这些研究的结果有可能改变我们对人类如何了解和浏览他们的社交世界的知识。这对于促进我们对影响本科生在STEM领域坚持不懈的社会因素的理解具有重要意义。该项目由理解神经和认知系统的综合策略(NCS)资助，NCS是一个多学科项目，由生物学(BIO)、计算机和信息科学与工程(CEISE)、教育和人力资源(EHR)、工程(ENG)以及社会、行为和经济科学(SBE)的主管部门联合支持。这项工作旨在为在真实的社会背景下理解人类学习提供基础。即将到来的本科生班级(由尚未建立大学社交网络的人组成)提供了一个独特的试验台，用于跟踪社交网络的出现及其在学生第一学年的过程中不断变化的配置。主要的假设是，个人对其社交网络的认知地图使他们能够更熟练地在他们的网络中导航。这项研究将通过功能磁共振成像(FMRI)的方式探索这种认知结构的神经表征，重点放在海马体和眶前叶/内嗅觉皮质)。将建立不同类别的计算模型(即基于状态转移的、基于后续表征的和基于潜在原因的)来测试社会网络学习的相互竞争的账户和社会认知地图的格式。这些模型将探索个人层面的参数，以评估个人的变异性是否会偏向这一学习过程，包括个人在社区中的位置，以及他们收集信息的能力，是否塑造了他们构建社会环境认知地图的能力。通过利用横截面和纵向设计，这项研究将提供一个组织框架，可以识别大脑如何代表有关复杂环境的社会知识，同时精确建模社交网络的认知地图如何实现有效的社交导航。这项工作将提供一个窗口，了解在现实世界中展开的社会学习过程背后的神经生物学机制。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Human life plays out in the vast landscape of our social networks. Successful social navigation can help buffer against loneliness and negative interactions, but it requires learning who is connected to whom: the latent structure underlying social relationships (e.g., cliques or hubs). The goal of this project, led by a team of researchers at Brown University, is to reveal the behavioral and neural signatures of how people learn and reason about real-world social networks. Despite the importance of understanding how people learn relational social knowledge, much remains unknown about how the brain organizes this knowledge. Spatial cognition research offers a window into this problem: A long line of animal research has demonstrated that the brain represents information about physical space in a ‘cognitive map’ that binds information about entities in the world and their relationships. Cognitive maps have been shown to speed learning in new or changing environments. This project tests whether, in mentally navigating their social networks, people recruit cognitive map-like representations similar to those used to navigate external physical space. By integrating advanced human imaging methods, computational modeling, social network science, and longitudinal sampling, the investigators will study a large cohort of first-year undergraduate students as they develop new friendships over the course of an academic year to investigate the behavioral and neural signatures of emerging knowledge about real-world social networks. The investigators will identify the neural representations of social information in this complex and dynamic environment. The results of these studies have the potential to transform our knowledge of how humans learn about and navigate through their social world. It has implications for advancing our understanding of social factors that contribute to the persistence of undergraduates in STEM fields. This project is funded by Integrative Strategies for Understanding Neural and Cognitive Systems (NCS), a multidisciplinary program jointly supported by the Directorates for Biology (BIO), Computer and Information Science and Engineering (CISE), Education and Human Resources (EHR), Engineering (ENG), and Social, Behavioral, and Economic Sciences (SBE). This work aims to provide a foundation for understanding human learning in real social contexts. An incoming undergraduate class (comprised of individuals who have yet to establish their college social network) provides a unique testbed to track the emergence of a social network and its shifting configuration over the course of the students’ first academic year. The main hypothesis is that an individual’s cognitive map of their social network enables them to navigate more adeptly through their network. The study will probe neural representations of this cognitive structure by means of function magnetic resonance imaging (fMRI) focusing on the hippocampus and orbitofrontal/entorhinal cortex). Different classes of computational models (i.e., state transition-based, successor representation-based, and latent-cause based) will be built to test competing accounts of social network learning and the format of social cognitive maps. These models will probe for person-level parameters to assess whether individual variability biases this learning process, including whether an individual’s position in their community, and thus their capacity to gather information, shapes their ability to build cognitive maps of their social environment. By harnessing both a cross-sectional and longitudinal design, the research will provide an organizing framework that can identify how the brain represents social knowledge about complex environments while precisely modelling how cognitive maps of a social network enable efficient social navigation. This work will provide a window into the neurobiological mechanisms underlying the social learning processes that unfold in the real world.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.