CAREER: Leveraging Online Behavior to Support Knowledge and Memory

这项研究将考察两个相关问题：(1)如何激励人们为开源软件和维基百科等生产公共产品的社区做出更多贡献；(2)如何通过使用他们在网上创建的支持回忆的内容来加强人们的自我概念和与他人的关系。这项工作将计算和社会科学深度交织在一起，使用对人们动机、目标和行为的洞察来驱动模型、算法和界面，这些模型、算法和界面利用人们的在线活动为个人和社会创造价值。这项活动的在线性质使其可以聚合为大型数据集，用于建模(例如，社交网络分析)和挖掘(例如，协作过滤)；研究的一个主要主题是有效地从人们已经进行的活动中挖掘出更多价值。理解人们在线行为的原因将导致建立过程模型，解释人们通过他们的行动产生的数据的重要特征，以及利用这些数据的新算法。例如，这项研究将在维基百科上模拟人们所采用的关键事件和角色如何随着时间的推移影响人们的贡献。这样的模型模型将驱动算法，在模型建议的上下文中将人们暴露给其他人、组、工具、政策和组规范，这将增加人们做出贡献的动机。理解用户的目标还将导致新的数据应用程序和更有效的呈现数据的界面。这项研究将通过一系列提示记忆的轻量级原型，以及通过分析社交媒体上的在线行为，来研究人们如何以及为什么会回忆。这项工作将产生从社交媒体活动中捕获记忆负载内容的算法，以及有效利用这些内容来支持回忆的界面。初步工作表明，自发、移动地发送适当选择的提醒有望增加人们从他们创建的内容中获得的价值。更广泛地说，设计这些模型、算法和界面的过程将导致在设计中使用社会科学理论的洞察力，以便捕获并与从业者分享，分析复杂的社交数据的新方法，以及产生有用的行为数据集，从而使其他研究人员受益。越来越多地参与像维基百科这样的公共产品，将改善成员的个人体验和他们创造的社会产品。在回忆领域开发的工具有可能改善许多人的生活，特别是随着人口老龄化。该教育计划通过参加会议和与其他实验室的暑期交流计划提供更丰富的研究经验。它还通过从技术和社会角度考察真实系统及其提供的数据的课程，帮助学生发展这项工作所需的跨学科态度和技能。

This research will examine two related issues: (1) how to motivate people to contribute more to communities such as open source software and Wikipedia that produce public goods, and (2) how to strengthen people's self-concepts and relationships with others by using content they create online to support reminiscence. This work deeply intertwines computing and social science, using insight about people's motivation, goals, and behavior to drive models, algorithms, and interfaces that leverage people's online activity to create value for individuals and society. The online nature of this activity allows it to be aggregated into large data sets for modeling (e.g., social network analysis) and mining (e.g., collaborative filtering); a major theme of the research is to effectively wring more value out of the activities people already do.Understanding why people act online will lead to process models that explain important features of the data people generate through their actions as well as new algorithms for exploiting that data. For example, the research will model how critical events and roles people adopt affect people's contributions over time in Wikipedia. Such models models will drive algorithms that expose people to other people, groups, tools, policies, and group norms in contexts the models suggest will increase people's motivation to contribute.Understanding users' goals will also lead to new applications for data and more effective interfaces for presenting it. The research will study how and why people reminisce through a series of lightweight prototypes that cue memories, as well as through analysis of online behavior in social media. This work will lead to algorithms that capture memory-laden content from activity in social media and interfaces that effectively use that content to support reminiscence. Preliminary work suggests that spontaneous, mobile delivery of appropriately chosen reminders promises to increase the value people derive from the content they create.More broadly, the process of designing these models, algorithms, and interfaces will lead to insights about using social science theory in design that can be captured and shared with practitioners, new methodologies for analyzing complex social data, and the production of useful behavioral datasets that will benefit other researchers. Increasing participation in public goods like Wikipedia will improve the individual experience of members and the social goods they create. Tools developed in the domain of reminiscing have the potential to improve many people's lives, especially as the population ages. The education plan provides for richer research experiences through conference attendance and summer exchange programs with other labs. It also helps students develop the interdisciplinary attitudes and skills needed for this work through courses that look at real systems and the data they provide from both technological and social perspectives.