东北大学建校100周年
工商管理学院系列学术报告第33场
报告题目：You Are What You Post: Determining Fairness Beliefs using Machine Learning
报 告 人：香港中文大学 张成龙 助理教授（副研究员）
报告地点：文管学馆B306
报告摘要：
Designing a system deemed “fair” to everyone is challenging because individuals typically hold different fairness beliefs. To address this challenge and design a fair system would require accessing data on a population's fairness beliefs and then further accommodating them. In this paper, we propose a framework to extract population fairness beliefs from social media comments on topics involving different fairness criteria. The framework includes both the approaches for constructing fairness beliefs and the tools for testing the validity of the constructed fairness beliefs. We demonstrate the model's validity using collected data from popular social media platforms and present additional interesting findings derived from the data experimentation. We find that no existing metrics in Natural Language Processing (NLP) can serve as effective features to predict fairness beliefs and that new features for learning fairness beliefs from the text comments are needed. Some other empirical findings concern the prediction power of unsupervised learning methods and the interpretation of cluster centroids. Our academic contributions are on the design of constructing and validating tools. In addition, the framework has significant managerial implications and can be used to resolve many real problems.
报告人简介：
张成龙，2021年毕业于德州大学达拉斯分校（UTD），现就职于香港中文大学（深圳）管理与经济学院，担任信息系统领域助理教授（副研究员）。主要研究方向为共享经济，区块链，人工智能和公平决策。其论文发表在《Management Science》等信息系统领域国际顶级期刊。同时，其担任《Management Science》，《Information Systems Research》，《MIS Quarterly》，《Production and Operations Management》等期刊的审稿人。此外，其作为主要参与人参与国家自然科学基金重点项目1项。
欢迎有兴趣的师生积极参加！