题　目：A Man with a Machine: Human–AI Augmentation for Sentiment Extraction from Firm Disclosure

嘉　宾：Jiexin Zheng, PhD candidate, the Hong Kong University of Science and Technology

主持人：魏煊  助理教授  tyc1286太阳成集团

时　间：2023年11月22日（周三）13:30-15:00

地　点：tyc1286太阳成集团 徐汇校区新上院N314室

内容简介:

We develop an augmentation approach that links artificial intelligence (AI) to human knowledge to improve sentiment extraction from firm disclosures. This approach uses the Loughran and McDonald (LM) dictionary to represent human knowledge and the word embedding model to adjust the sentiment strength of words over time. We evaluate our approach using firms’ 10-K reports and find that the augmented sentiment measures consistently explain abnormal market returns immediately after the release of 10-K reports. This indicates that inconsistent results in the literature are driven by inaccurate sentiment measures. We also find that our augmented sentiment measures, when used as a word reweighting scheme for machine learning, have improved economic significance and explanatory power for abnormal market returns. Next, we find that our augmented sentiment measures help predict firms’ future earnings. Our research provides evidence suggesting that the proposed augmentation method mitigates the impact of managers’ strategic manipulation on disclosure sentiment. Further analyses demonstrate the robustness and generalizability of our method, and reveal that our augmented sentiment measures outperform conventional NLP and BERT-related models in both explaining and predicting market reactions and future earnings performance. Our work highlights the complementarity between human knowledge and AI in the context of textual information processing in financial markets.

演讲人简介：

Jiexin Zheng is a Ph.D. candidate in the Department of Information Systems, Business Statistics, and Operations Management at the Hong Kong University of Science and Technology. He works on multidisciplinary research that employs various techniques, such as machine learning methods, experimentation, surveys, and econometric modeling, to address important business problems and provide multifaceted insights. His primary research focus is on human-AI interaction. He has conducted research and developed methodologies that leverage the complementary strengths of human intelligence and artificial intelligence to enhance the processing of large-scale data. Additionally, he explores strategies for maximizing the benefits of AI while minimizing its risks. His research has been accepted by the Journal of Management Information Systems and the Management and Organization Review, and has been presented at leading Information Systems conferences.

欢迎广大师生参加！