第五届网络经济博弈论坛将于2025年11月14日至16日在吉林大学举办。本次会议由中国计算机学会（CCF）主办，CCF计算经济学专委会、吉林大学数量经济研究中心承办。论坛采取线下的方式进行，是网络、计算经济学及博弈论等相关主题的研讨会，聚焦于算法博弈论、信息与计算社会科学、互联网经济学、网络博弈等相关的前沿研究。

会议内容

会议内容包括主旨报告、分论坛报告、墙报展示等环节。届时，我们将邀请国内外计算经济学领域的知名专家进行学术演讲，并围绕计算经济学的前沿进展展开深入研讨，以促进国内外学者之间的学术交流。我们热烈欢迎国内外计算经济学及相关领域的学者莅临本次盛会，进行学术交流与探讨。

本次推送介绍本次论坛的五个分论坛和十九场报告。

参会事宜

举办时间：2025.11.14 - 11.16

报名时间：即日至2025.11.16

地点：吉林省长春市吉林大学前卫南区东荣大厦

联系方式：

谢纪新  电话：15948315805

汪子琪  电话：15204357731

注册方式

会议门票：

报名链接：https://ccf.org.cn/tcce202502

或扫描下方二维码报名：

会议日程

数字经济分论坛

（一）李张体（中国联通集团）

李张体，正高级工程师，中国联通集团首批领域专家，现任中国联通软件研究院数据中台研发事业部总经理。深耕IT、DT领域20余年，牵头国资委GJGC、“百大工程”、大数据原创技术策源地、工信部人工智能产业链链主、国家自然科学基金项目等国家重点工程等8项国家级重点工程。主导建成全球行业规模最大、最先进的IT核心支撑系统cBSS，打造新型数据基础设施，构建国家数据产业底座。个人荣获国资委青年文明号、中国联通集团工匠等10余项荣誉称号，研究成果荣获中国电子学会、数博会等多项省部级科技奖项，本人授权受理专利9项，牵头制定数据集、数据基础设施国家标准5项，发表SCI等高水平论文3篇。

Title：中国联通高质量数据集关键技术研究与应用实践

Abstract：数据作为国家确立的第五大数据要素，已成为驱动大模型训练与推理过程的基础性资源，其高质量、规模化与多样性特征对人工智能系统的演进具有重要支撑作用。在人工智能发展新阶段，多模态数据融合作为推动模型架构跃迁与产业体系革新的关键路径，正逐步成为构建智能系统核心竞争力的基础资源。中国联通坚持“融合创新”发展战略，历经信息化、数字化与智能化三个阶段，持续打造数智新基建、数智新技术、数智新应用，积极落实“数据要素×”与“人工智能+”国家行动，承担工信部人工智能产业链“锻长板”任务，攻关多模态数据融合、智能标注与质量评估等关键技术，首创“运营商链主+生态协同”的数据集建设模式，联合产学研用多方力量，共同推进行业高质量数据集的构建与运营。在此基础上，中国联通创新提出高质量数据集“三个一”体系，系统推进AI时代数据治理体系建设，实现从传统数据平台向AI多模态数据平台的智能化转型升级。通过构建“采、洗、标、测、用、评”一站式数据集生产流水线，持续促进数据要素价值的高效转化，有效支撑大模型在精度提升与迭代加速方面的能力建设，并在政务、文化、工业等重点领域实现了从标杆项目突破到规模化应用的系统推进，为数字中国建设注入了持续的数据要素动能，系统构建了从“数治”到“数智”的产业革新路径。

（二）盖志强（智唐科技）

盖志强，现任智唐科技首席数据官（CDO），具有15年以上企业信息化以及互联网工作经验，在供应链金融、智能制造及建筑数字化领域具有丰富数字化转型规划、落地经验以及大数据规划与数据中台产品设计能力，同时具有多年数仓建设/数据中台建设及数据全生命周期治理、数据指标体系、数据标签/用户画像系统的规划设计经验。先后负责大型智慧园区，智慧城市，智能制造和大数据平台等项目交付。先后就职于神州数码，泛微软件，便利蜂，苏宁集团，京东集团，同仁堂集团等。获得工业信息化部智能制造IT规划师，纳入工信部专家库专家；北京市数字乡村建设领军人才；DAMA专家会员。

Title：数据智能在数字经济领域的应用探讨

Abstract：数据智能是数字经济的核心引擎，深度重构产业价值与发展模式。围绕数字产业化、产业数字化两大主线，聚焦智能决策、效率提升、创新赋能三大关键方向，呈现数据智能在产业链协同、精准服务、风险管控等场景的落地应用。通过数据整合与算法赋能，其助力打破数据壁垒、优化资源配置，驱动产业升级与模式创新。

（三）刘璇（北京大学）

刘璇，北京大学经济学院博士后，理学博士。博士期间，曾主持中国人民大学研究生科学研究基金项目1项，参与中国人民大学“交叉创新研究计划”项目2项，多次荣获中国人民大学学业奖学金一等奖、学习优秀奖学金，并获评“优秀毕业生”“优秀学生干部”“三好学生”等荣誉称号。目前正参与国家自然科学基金面上项目1项。在《IEEE Transactions on Computational Social Systems》等国际权威期刊发表论文，并在ACM International Conference on Computing Frontiers等国际学术会议发表研究成果。研究方向聚焦区块链与数据要素流通机制设计、多方拍卖与定价机制、数据市场架构设计等领域。

Title：基于消费者画像的数据定价机制研究

Abstract：随着数字经济发展，数据已成为具有显著经济价值的新型生产要素，催生了全球数据要素市场的蓬勃兴起。然而，现行市场机制呈现出显著的结构性缺陷：现有数据交易所普遍侧重于供给侧的资质审核、合规监管与数据质量评估，而对需求侧的消费者画像与信用资质评估机制存在系统性缺失。这种供需评估体系的非对称性引发了市场信任机制的失灵，数据提供者因担忧优质数据资产流向低信用或恶意消费者，而产生强烈的市场退出倾向，进而导致数据滥用风险加剧、市场稳定性受损以及长期收益预期恶化等一系列问题。针对上述市场失灵现象，本研究创新性地将消费者画像变量纳入数据定价的理论框架，构建了基于消费者画像与保留价格双重约束的数据定价机制，旨在通过机制设计实现对低画像消费者的有效甄别与排除，从而保障数据提供者的长期收益稳定性。研究首先在理论层面严格证明了基于画像的广义第二价格（GSP）拍卖模型在不完全信息博弈环境下贝叶斯纳什均衡的存在性与有效性，为机制的理论可靠性提供了坚实基础。其次，针对具有异质性画像特征的消费者群体，本研究推导出差异化的最优个性化保留价格函数，实现了定价策略的精准化配置。进而，系统性地分析了消费者画像机制的引入及位置系数的动态调整对数据提供者收益函数的影响机理与作用路径。最后，通过构建系列计算实验，本研究实证检验了所提出定价机制在最大化数据提供者长期收益方面的有效性与稳健性，为数据要素市场的机制优化提供了理论依据与实践参考。

图片
数字资产安全分论坛

（一）王也（澳门大学）

王也，澳门大学电脑及资讯科学系、金融及商业经济系双聘助理教授，博导。在苏黎世联邦理工学院获得博士和硕士学位，在北京大学获得学士学位。主要从事信息安全、人机交互、金融科技等方面的研究工作。近年来在JFE、S&P、USENIX Security、CCS、NDSS、CHI、CSCW、WWW、WINE、IJCAI等国际期刊与会议发表学术论文30余篇。主持/共同主持国家科技部重点研发计划、澳门科学技术发展基金面上项目等科研项目。

Title：Nudge Interventions for Address Poisoning Phishing

Abstract：Address poisoning phishing is a form of phishing in Web3 that deceives users into transferring funds to phishing addresses by exploiting their misinterpretation of wallet interface information. While prior research has largely focused on real-time detection, little is known about prevention from the user perspective. Drawing inspiration from studies on nudges in traditional phishing, we designed targeted interventions at three critical stages of the phishing flow. We evaluated the effectiveness of these designs through a between-subjects online experiment (n = 281) and complemented the findings with semi-structured interviews with 20 participants to gain deeper insights into user behavior. Our results reveal substantial variation in the effectiveness of interventions across different stages, and we discuss the unique challenges of designing nudges for address poisoning phishing compared to traditional phishing, as well as how to intervene effectively in long behavioral chains where users' intended actions must be redirected for protection.

（二）Leon Witt（上海数学与交叉学科研究院）

Dr. Leon Witt is a Post-Doc at the Shanghai Institute for Mathematical and Interdisciplinary Sciences (SIMIS), specializing in decentralized and incentivized artificial intelligence. He has worked in strategy consulting at McKinsey’s Digital Capability Center in Germany, as a software engineer in Silicon Valley, and led the decentralized AI research group at the Fraunhofer Heinrich Hertz Institute in Berlin. He earned his Ph.D. from Tsinghua University, where he became the first German to graduate from the Department of Computer Science. His research focuses on blockchain-based incentive mechanisms, multi-task peer prediction, and federated learning to build transparent and trustworthy AI ecosystems.

Title：Knowledge Free Correlated Agreement to Incentivize Federated Learning

Abstract：In Federated Learning (FL), accurately assessing client contributions is a challenge. Explicit methods like Shapley value (SV) require considerable computational resources, while implicit peer-prediction methods such as Correlated Agreement (CA) necessitate knowledge about reported signals and are vulnerable to attacks. To overcome these hurdles, we introduce Knowledge-Free Correlated Agreement (KFCA), an efficient multi-task peer prediction mechanism for FL. KFCA enhances contribution assessment and resolves CA's label-flipping issue without needing knowledge about underlying distributions. We provide theoretical evidence that under categorical data conditions, a client's optimal strategy is to train the model and report honestly, leading to a robust equilibrium. Experimental validation of KFCA against various CA and Shapley value estimates, and experimental results on data from a real-world application of Printed Circuit Board (PCB) quality inspection, demonstrate its practical efficacy, making KFCA a potential candidate for integration into blockchain-based smart contracts, promoting decentralized FL.

（三）曹露（复旦大学）

曹露，复旦大学与上海数学与交叉学科研究院双聘助理教授，悉尼大学应用数学博士。主要研究方向包括平均场博弈、机制设计以及数学与交叉学科研究。曾任悉尼大学讲师、北京大学北京国际数学研究中心博士后研究员。此前担任悉尼某投资集团研究总监，负责研究报告撰写与投资策略开发，具有丰富的跨学科科研与产业研究经验。

Title：AI驱动的数字资产安全与隐私保护：智能合约、共识机制与隐私计算协议的安全分析

Abstract：随着数字资产与区块链技术的广泛应用，其安全性与隐私性问题成为数字经济发展的关键挑战。本报告提出了一种AI驱动的数字资产安全与隐私评估框架，通过智能化分析手段实现对智能合约、共识机制和隐私计算协议的系统化安全评估与验证。首先，在智能合约安全性分析方面，引入深度学习与符号执行相结合的检测模型，提高漏洞识别的自动化与准确性。其次，在区块链共识机制安全评估方面，基于博弈论与强化学习构建动态安全模型，量化并优化共识机制在去中心化环境下的稳定性与抗攻击能力，重点防范自私挖矿与多数攻击等风险。再次，在隐私计算协议安全验证方面，结合机器定理证明与密码学方法，提出可自动化验证协议正确性与安全边界的分析框架。通过上述三类技术的协同融合，本研究形成了一套可扩展、可验证的数字资产安全与隐私评估体系，为数字要素市场提供自动化安全评估、隐私保护与合规性检测的新途径，助力数字经济基础设施的安全与可信发展。

算法博弈论分论坛

（一）冯逸丁（香港科技大学）

冯逸丁，香港科技大学工业工程及决策分析系助理教授。在加入香港科技大学之前，他曾先后在芝加哥大学布斯商学院及微软研究院新英格兰分部担任博士后研究员。他于2021年获得美国西北大学计算机科学博士学位，并于2016年毕业于上海交通大学ACM班，获学士学位。他的研究兴趣位于运筹学、经济学与计算，以及理论计算机科学的交叉领域。相关研究成果已发表在 Management Science、Operations Research 等国际顶级期刊，以及STOC、FOCS、SODA、EC、ITCS、WINE 等理论计算机科学与计算经济学重要会议上。他曾获得INFORMS Auctions and Market Design Michael H. Rothkopf青年学者论文奖，以及APORS青年学者最佳论文奖。

Title：Competition Complexity in Multi-Item Auctions: Beyond VCG and Regularity

Abstract：We quantify the value of the monopoly's bargaining power in terms of competition complexity—that is, the number of additional bidders the monopoly must attract in simple auctions to match the expected revenue of the optimal mechanisms—within the setting of multi-item auctions. We show that for simple auctions that sell items separately, the competition complexity is Θ(n/α) in an environment with n original bidders under the slightly stronger assumption of α-strong regularity, in contrast to the standard regularity assumption in the literature, which requires Ω(n·ln(m/n)) additional bidders. This significantly reduces the value of learning the distribution to design the optimal mechanisms, especially in large markets with many items for sale. For simple auctions that sell items as a grand bundle, we establish a constant competition complexity bound in a single-agent environment when the number of items is small or when the value distribution has a monotone hazard rate. Some of our competition complexity results also hold when we compete against the first-best benchmark (i.e., optimal social welfare).

（二）伏虎（上海财经大学）

Hu Fu is an associate professor in the Institute for Theoretical Computer Science at Shanghai University of Finance and Economics. The institute is part of the School of Computing and Artificial Intelligence. I received my PhD from the Department of Computer Science at Cornell University, under the supervision of Professor Robert Kleinberg. I was then a postdoc at Microsoft Research, New England Lab, and later at Caltech, Computing and Mathematical Sciences. I was assistant professor in the Department of Computer Science of the University of British Columbia from 2016 to 2020. My research takes a computational perspective on problems arising from economic contexts. I am also interested in theoretical computer science in general.

Title：Price Stability and Improved Buyer Utility with Presentation Design: A Theoretical Study of the Amazon Buy Box

Abstract：Platforms design the form of presentation by which sellers are shown to the buyers. This design not only shapes the buyers' experience but also leads to different market equilibria or dynamics. One component in this design is through the platform's mediation of the search frictions experienced by the buyers for different sellers. We take a model of monopolistic competition and show that, on one hand, when all sellers have the same inspection costs, the market sees no stable price since the sellers always have incentives to undercut each other, and, on the other hand, the platform may stabilize the price by giving prominence to one seller chosen by a carefully designed mechanism. This calls to mind Amazon's Buy Box. We study natural mechanisms for choosing the prominent seller, characterize the range of equilibrium prices implementable by them, and find that in certain scenarios the buyers' surplus improves as the search friction increases.

（三）刘圣鑫（哈尔滨工业大学〈深圳〉）

刘圣鑫，现任哈尔滨工业大学（深圳）计算机科学与技术学院副教授。他于香港城市大学获得博士学位，随后于新加坡南洋理工大学从事博士后研究工作。他的研究兴趣为计算经济学和理论计算机科学。近期，他主要从事资源公平分配问题的研究。所发表论文荣获AAAI会议最佳学生论文奖和FAW会议最佳论文奖。

Title：Approximability Landscape of Welfare Maximization within Fair Allocations

Abstract：The problem of fair allocation of indivisible goods studies allocating a set of $m$ goods among $n$ agents in a fair manner. While fairness is a fundamental requirement in many real-world applications, it often conflicts with (economic) efficiency. This raises a natural and important question: How can we identify the most welfare-efficient allocation among all fair allocations? This paper gives an answer from the perspective of computational complexity. Specifically, we study the problem of maximizing utilitarian social welfare (the sum of agents' utilities) under two widely studied fairness criteria: envy-freeness up to any item (EFX) and envy-freeness up to one item (EF1). We examine both normalized and unnormalized valuations, where normalized valuations require that each agent's total utility for all items is identical. The key contributions of this paper can be summarized as follows: (i) we sketch the complete complexity landscape of welfare maximization subject to fair allocation constraints; and (ii) we provide interesting bounds on the price of fairness for both EFX and EF1.

This talk is based on joint work with Xiaolin Bu, Zihao Li, Jiaxin Song, and Biaoshuai Tao, which is published in EC-2025.

（四）陶表帅（上海交通大学）

Biaoshuai Tao is a (tenure-track) associate professor at John Hopcroft Center for Computer Science at Shanghai Jiao Tong University. Biaoshuai Tao joined Shanghai Jiao Tong University since 2020. In 2020, he received his Ph.D. degree in computer science at the University of Michigan, Ann Arbor. His research interests mainly include the interdisciplinary area between theoretical computer science and economics, including computational social choice, resource allocation problems, social network analyses, and algorithmic game theory in general. Before joining the University of Michigan, Biaoshuai was employed as a project officer at Nanyang Technological University in Singapore from 2012 to 2015, and he received the B.S. degree in mathematical science with a minor in computing from Nanyang Technological University in 2012.

Title：Likelihood of the Existence of Average Justified Representation

Abstract：We study the approval-based multi-winner election problem where n voters jointly decide a committee of k winners from m candidates. We focus on the axiom \emph{average justified representation} (AJR) proposed by Fernandez, Elkind, Lackner, Garcia, Arias-Fisteus, Basanta-Val, and Skowron (2017). AJR postulates that every group of voters with a common preference should be sufficiently represented in that their average satisfaction should be no less than their Hare quota. Formally, for every group of l*n/k voters with l common approved candidates, the average number of approved winners for this group should be at least l. It is well-known that a winning committee satisfying AJR is not guaranteed to exist for all multi-winner election instances. In this paper, we study the likelihood of the existence of AJR under the Erdős-Rényi model. We consider the Erdős-Rényi model parameterized by p\in[0,1] that samples multi-winner election instances from the distribution where each voter approves each candidate with probability p (and the events that voters approve candidates are independent), and we provide a clean and complete characterization of the existence of AJR committees in the case where m is a constant and n tends to infinity. We show that there are two phase transition points p1 and p2 (with p1<=p2) for the parameter p such that: 1) when pp2, an AJR committee exists with probability 1-o(1), 2) when p1<p</p

This is a joint work with Qishen Han, Lirong Xia, Chengkai Zhang, and Houyu Zhou. This work has been accepted by SODA’26.

社会经济分论坛

（一）刘洋（吉林大学）

刘洋，数量经济学博士，吉林大学商学与管理学院应用金融系副教授，硕士生导师。科研工作聚焦于宏观经济与数量经济学方法领域，近期聚焦于与计算经济学，特别是电子工程经济学研究方向。已完成的科研项目包括交付给国家信息中心，中国人民银行的数量经济学方法相关的系统建设课题。在《经济评论》、《管理科学》、《系统工程理论与实践》等期刊发表多篇论文，主讲《人工智能与金融应用》等金融科技专业课程，出版《贝叶斯计量经济学前沿理论及应用》等专著两部。在贝叶斯计量经济学和异质性代理人宏观经济模型领域已开发和提交多项软件和发明专利，其科研团队研发的系列数量经济学芯片IP服务于定量遥感、节能优化等金融科技相关领域。

Title：Econometrichip：数量经济学EPU芯片设计与应用

Abstract：随着人工智能技术的发展与应用范围的扩大，不同学科对新型算力的需求也呈现出多样化和专业化的发展趋势。动态规划等数量经济学核心算法的算力瓶颈已成为宏观经济理论模型研究的突出障碍。相比与深度神经网络等更适用于自然科学的GPU算力发展，侧重于社会行为模拟的计算经济学基础设施明显薄弱。本文基于计算经济学领域在电子工程经济学方向上的近期突破，提出全新的算法改进与芯片设计优化，实现Econometric Processing Unit（EPU）的计算架构，为数量经济学提供专属新型算力。本文的研究表明，EPU在动态规划优化求解方面，相较于现有CPU和GPU架构具有明显优势。本文提出的社会经济行为模拟的新型算力基础设施，提升经济学理论研究的模拟计算能力进入全新阶段。

图片
（二）孟佶贤（北京林业大学）

孟佶贤，北京林业大学经济管理学院副教授，研究生导师，经济学博士，管理科学与工程专业博士后。博士期间，获得CSC资助在荷兰马斯特里赫特大学博士公派联合培养1年。曾主持国家自然科学基金青年项目1项，中国博士后科学基金面上项目1项，中央高校基本科研业务费项目1项。目前正主持北京市自然科学基金面上项目1项，中央高校基本科研业务费项目1项。参与教育部人文社科重点研究基地重大项目1项，国家自然科学基金面上项目3项，国家发展改革委和国家林草局委托课题4项。在《中国社会科学》《中国管理科学》《南开管理评论》《保险研究》《Journal of Graph Theory》《Journal of Systems Science and Complexity》等期刊发表论文25篇。担任北京林业大学学报（社会科学版）首届青年编委；中国优选法统筹法与经济数学研究会网络科学分会理事；中国计算机学会计算经济学专业委员会执行委员。

Title：多模态融合情绪线索与资本市场反应

Abstract：本研究基于多模态分析框架，构建了管理者团队路演多模态融合积极情绪指标，并探究其对企业IPO收益表现的影响。研究发现，管理者团队在多模态层面展现的积极情绪显著提升了IPO短期收益。模态拆解研究表明，相比语言模态，路演视听模态的积极情绪表现对IPO短期收益的影响更显著。机制分析发现，管理者团队路演多模态积极情绪表现对投资者情绪的拉动作用是促进IPO短期收益上涨的重要原因。调节效应研究表明，较高的主持人面部吸引力、管理者半面对称性以及保荐代表人面部尽职感知强化了路演视觉模态积极情绪对IPO短期收益的正向影响。与此同时，高管发音标准和流畅程度强化了听觉模态积极情绪的影响。进一步分析发现，管理者团队多模态情绪表现积极的企业未来更易出现收益反转和股价波动加剧，而经营绩效并未改善。分组异质性检验结果表明，本文的主要研究发现在上市前信息不对称程度高、承销商声誉良好以及风险投资持股比高的企业分组中更显著。本研究提出了一套适用于非结构化视频数据的多模态特征提取框架，通过模态融合、拆解以及多源异质线索的提取等，系统揭示了多模态线索对企业资本市场反应的影响。

（三）张国权（吉林大学）

张国权，吉林大学商学与管理学院教授，博士生导师。主要从事行为运筹与决策科学、物流与供应链管理、碳减排与可持续发展等方面的研究。2014年入选吉林大学优秀青年教师重点培养计划。先后主持国家社会科学基金、国家自然科学基金、教育部人文社科基金、吉林省社会科学基金、吉林省科技发展计划、中国博士后基金等多个项目。第一作者或通讯作者在EJOR、Omega、IJPR、ESWA等学术期刊上发表论文多篇，为多个 SSCI/SCI 学术期刊审稿人。

Title：Probabilistic Selling: A Game-Theoretic Approach to Tackling Near-Expired Food Waste

Abstract：Retailers commonly employ price discounts and donations to address the issue of near-expired food waste. However, price discounts often lead to consumer fatigue, while donations may create conflicts of interest between retailers and charitable organizations. In response to these challenges, this study investigates the strategy of near-expired surprise bags—such as blind boxes—as an alternative approach. It develops an analytical framework for the sale of near-expired surprise bags to explore their optimal design and pricing.

互联网经济分论坛

（一）刘正阳（北京理工大学）

刘正阳，北京理工大学计算机学院长聘副教授，特聘研究员，博士生导师，研究方向聚焦于博弈论与计算复杂性。已在STOC、WWW、AAAI、AAMAS等顶级会议与期刊发表论文20余篇，并先后主持国家自然科学基金青年项目与面上项目。现任中国计算机学会（CCF）理论计算机科学专委会、计算经济学专委会执行委员，以及中国运筹学会博弈论分会理事。

Title：一种基于排序的大模型评估方法

Abstract：我们针对传统大模型评估方法中依赖固定范式、难以衡量其开放性与主观性的局限，探索博弈论技术在评估中的应用。我们提出“自动互评估”机制，通过模型间的博弈与互评，对比人类投票行为，以检验其与人类判断的一致性。结合博弈论投票算法整合模型互评，本研究首次将互评估、博弈论聚合与人类验证相融合，探讨了理论预测与人类评估间的异同，为大模型评估方法提出了新的尝试。

（二）余皓然（北京理工大学）

Haoran Yu is currently an Associate Professor with the School of Computer Science and Technology, Beijing Institute of Technology. He received the Ph.D. degree from the Department of Information Engineering, the Chinese University of Hong Kong in 2016. From 2015 to 2016, he was a Visiting Student with the Yale Institute for Network Science and the Department of Electrical Engineering, Yale University. From 2018 to 2019, he was a Post-Doctoral Fellow with the Department of Electrical and Computer Engineering, Northwestern University. His current research interests lie in the interdisciplinary area between game theory and artificial intelligence, with a focus on real human strategic behavior analysis. His work has been published mainly in conferences and journals within the fields of artificial intelligence and networking (e.g., AAAI, IJCAI, ACM SIGMETRICS, ACM MobiHoc, IEEE INFOCOM, IEEE/ACM TON, IEEE JSAC, and IEEE TMC).

Title：Modeling Human Strategic Behavior in the Era of LLMs

Abstract：Modeling strategic human behavior has long been a central research problem. Traditionally, work in this area has relied on hand-crafted models developed for relatively simple games and settings. As interactions grow more complex and feature spaces expand, hand-crafted approaches become increasingly impractical. Large language models (LLMs) are reshaping how we develop models of human strategic behavior. Their broad knowledge and emergent reasoning capabilities enable them to propose, evaluate, and refine behavioral models more efficiently. This talk provides a high-level perspective on how LLMs can scale the discovery of interpretable behavioral models from observational data, and outlines the open research questions that arise in the era of LLMs.

（三）何翘楚（南方科技大学）

何翘楚现为南方科技大学商学院长聘副教授，清华大学工学学士、加州（伯克利）大学运筹学博士，国家特聘专家（青年），主要研究运营管理和运筹学在智慧城市和人智交互等领域的应用。在管理科学与工程及其相关领域顶级期刊和会议（MSOM，POM，TRB，Biometrika等）发表近50篇论文（其中UTD24种发表或返修近20篇）。主持国自然（NSFC）面上（两项）、深圳市科创委、香港基础研究基金（GRF）等项目。担任中国管理科学与工程学会理事，CCF计算经济学会执委，中国管理现代化研究会管理与决策科学专委，广东省机械工程学会物流工程分会副理事长。担任工程院院刊FEM、INFORMS会刊Service Science, MSOM/POM等UTD24期刊的编委或审稿人。曾任教于北卡州立UNC Charlotte和香港科大HKUST。课题组成员任教于华中科技大学、中科大商学院、深圳技术大学、英国利物浦大学管理学院等高校。

Title：Seeding Influencers with Marketing Goals and Limited Network Information

Abstract：In this paper, we address the influencer selection problem in influencer marketing within social networks. We consider an advertising agency that aims to assign influencers to multiple marketing campaigns to achieve predefined marketing goals while minimizing overall costs. The agency operates under the constraint of limited network information. To tackle this challenge, we propose a novel estimation approach for evaluating influencer engagement that leverages both rank information of influencers and distributional characteristics of the social network. In the first stage, we apply a rank aggregation method to estimate the first-level influence of each influencer and establish theoretical guarantees on the likelihood of recovering the true rank. In the second stage, we approximate the final network activations via a fixed-point equation, integrating this estimation into a robust optimization framework for downstream decision-making. Experimental results demonstrate that the proposed method achieves a substantial cost reduction compared to baseline approaches, while maintaining competitive marketing performance.

（四）沈蔚然（中国人民大学）

沈蔚然现任中国人民大学高瓴人工智能学院准聘副教授。他本科毕业于清华大学电子工程系，2019年于清华大学交叉信息研究院获博士学位，2019年至2020年于卡内基梅隆大学担任博士后研究员。主要研究方向为多智能体系统、博弈论、机制设计和机器学习，在相关领域国际会议及期刊发表高水平论文四十余篇，出版英文专著2本，中文教材2本。担任AAAI、WWW、IJCAI等多个国际会议的高级程序委员会委员、领域主席，以及AIJ、JET、TKDE等多个期刊的审稿人。现担任CCF理论计算机专委、计算经济学专委执行委员，中国运筹学会博弈论分会理事等职务。在机制设计方面的研究成果已在百度、字节跳动、快手等互联网平台落地实现。

Title：互联网广告中的优惠券机制设计与应用

Abstract：在互联网广告中，给用户发送优惠券可以提升广告的点击率和转化率，好的优惠券发放策略可以提升平台收益、社会福利，也可以降低用户购买成本。我们研究了优惠券的设计与发放机制，分析了优惠券机制的激励兼容条件，并根据该条件，提出了一种优惠券优化算法。我们在模拟和真实数据中进行了实验，验证了所提算法的有效性。

学生论坛

（一）马允轩（北京大学）

马允轩，北京大学前沿计算研究中心23级博士生，导师为邓小铁教授。主要研究领域为微观经济学问题的机器学习算法。

Title：Incentivizing Data Trading via Profit Reallocation

Abstract：Data is increasingly recognized as a valuable resource; yet, organizations are reluctant to share data across domains, often referred to as "data silos". The reason behind this phenomenon, as well as the remedy to resolve it, have been less understood. In this talk, we try to understand this phenomenon by introducing an incentive-based model where "data silos" is an economic consequence. We then show that "data silos" can be alleviated by introducing a "profit reallocation mechanism", which reallocates the profits of sellers to others.

（二）杨宗森（香港中文大学〈深圳〉）

Zongsen Yang is a PhD student at The Chinese University of Hong Kong, Shenzhen. He received his Bachelor's degree from the School of Management, Zhejiang University in 2022. His research has been published in leading journals and conferences such as Manufacturing & Service Operations Management and WINE. He was also a finalist for the INFORMS Service Science Best Student Paper Competition. His work has received media coverage in outlets such as Insurance Business and Financial Times (Chinese).

Title：Dynamic Competitive Learning and Imitation with Independent Demand Models

Abstract：In online markets crowded with similar sellers, pricing algorithms play a decisive role in shaping market competition. Yet firms often find it difficult to observe or track all competitors’ prices. Facing informational limits and rapid price adjustments, some firms may ignore competitors’ information and use simple independent demand models (IDMs) to predict demand and update prices, while others anchor their prices to selected competitors. Despite its prevalence, their dynamic interaction remains poorly understood. To this end, we establish a competitive learning framework where firms either update their beliefs via IDMs or imitate competitors’ prices. We prove that when all firms learn independently through IDMs, they overestimate market intercepts, and prices converge to the Nash equilibrium. However, as more firms adopt imitation, they underestimate price sensitivity, pushing prices toward perfect collusion. We further provide discrete and continuous characterizations of pricing correlation and prove that stronger (weaker, resp.) pricing correlation shifts market outcomes closer to collusive (competitive, resp.) prices. These results stand in sharp contrast to traditional static, full-information models of price competition and highlight how correlated learning behaviors shape equilibrium outcomes in dynamic competitive environments and offer new implications for detecting and regulating algorithmic collusion.

组织单位