Collaborative Research: RI: Medium: Informed, Fair, Efficient, and Incentive-Aware Group Decision Making

在我们这个快节奏和日益网络化的世界里，做出公平、公正和明智的决定比以往任何时候都更加重要，也更加困难。在一群人必须达成共识的情况下，在各种各样和迅速变化的信息面前，实现公平和公正变得更加困难。为此，本项目将结合社会选择理论和信息启发的研究，创建一个新的研究方向——知情群体决策。这个新的研究领域通过明确地考虑信息对代理最终决策的作用，扩展了当前群体决策的模型和机制。最终目标是开发新的模型和方法，可以用来激励个人，以确保群体决策达到预期的结果。这项研究承诺跨机构、教育和社会影响，并将扩大未被充分代表的群体在计算研究中的参与，培养高素质的专业人员，并吸引来自未被充分代表的群体的学生在计算相关领域进行研究。本研究包括基础研究的三个维度和理论与实践衔接的一个方向。维度1：表征旨在开发新的模型，以结合智能体的主观和客观偏好、信息和对查询的响应。维度2：聚合旨在为知情群体决策引入新的效率和公平标准，并设计新的机制来实现诚实、合作的主体。维度3：激励旨在通过提出新的均衡概念，对代理人行为进行分析，设计新的激励意识机制，解决代理人在知情群体决策中的激励问题。为了将理论与实践相结合，本项目开发的模型、算法和机制将通过各种教育和推广活动，在开源的在线偏好报告和聚合（OPRA）系统上进行部署、验证和完善。该奖项反映了美国国家科学基金会的法定使命，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。

In our fast-paced and increasingly online world making fair, equitable, and informed decisions is both more important and harder than ever. In situations where a group must come to a consensus, in the presence of varied and rapidly changing information, achieving fairness and equity becomes even more difficult. In response, this project will combine research from social choice theory and information elicitation to create a new research direction called informed group decision making. This new research area extends current models and mechanisms of group decision making by explicitly accounting for the role that information has on agents' final decisions. The final goal is to develop new models and methods that can be used to incentivize individuals to ensure group decisions achieve a desired outcome. This research promises cross-institutional, educational, and societal impacts and will broaden the participation of underrepresented groups in computing research, train highly qualified professionals, and engage students from underrepresented groups to pursue studies in computing-related fields.This research consists of three dimensions for foundational research and one direction for bridging theory and practice. Dimension 1: Representation aims to develop novel models for combining agents’ subjective and objective preferences, information, and responses to queries. Dimension 2: Aggregation aims to introduce novel efficiency and fairness criteria for informed group decision making, and design novel mechanisms to achieve them for truthful, cooperative agents. Dimension 3: Incentives aims to address agents’ incentives in informed group decision making by proposing novel equilibrium concepts, conducting analysis of agents’ behavior, and designing novel incentive-aware mechanisms. To bridge theory and practice, the models, algorithms, and mechanisms developed in this project will be deployed, validated, and refined at the open-source Online Preference Reporting and Aggregation (OPRA) system via various educational and outreach activities.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.