第一场
时间：2022年10月19日（周三）14：00
地点：腾讯会议324 386 049
报告形式：线上
报告人、报告题目及报告概要：
董直庆：工业智能化会助力实现碳达峰吗？约束条件与中国证据
摘要：破解全球极端天气频发方向之一就在于有效实现节能减排，在全球关注气候治理同期，工业智能化正全面赋能实体经济转向智能化生产。然而，现有文献既未有效识别一国工业化智能化水平，又未关注工业智能化能否助力节能减排。基于此，本文采用中国2011-2019年面板数据，采用熵权法估算工业智能化水平，检验工业智能化与节能减排能否同向发展，以及这类问题背后的形成机制和约束条件。结果表明，工业智能化并非一定存在节能减排效应，通常其与碳排放表现出显著“倒U”型关系且门槛值2.78，即工业智能化突破门槛值后才会降低碳排放。分组检验发现，不同情境工业智能化的碳减排效应差异显著，绝大部分地区工业智能化未正向降低碳排放，劳动和技术密集型行业工业智能化碳减排效应明显，但资本密集型行业则会反向提高碳排放。一般地，工业智能化会借助绿色技术与能源生产清洁化实现碳减排，但能源消费结构和要素配置机制却未出现预期效果。
侯仕军：企业共同富裕测度：基于中国上市公司2011-2020年的数据
摘要：我们立足共同富裕理论、政策及实践，从宏观到微观、跨层次地综合三次分配、利益相关者理论，建构和验证一套企业共同富裕测度指标体系。我们首先从利益相关者视角切入、基于层次分析法（AHP）构建一套企业共同富裕测度指标体系，并衍生性地划分企业共同富裕的双元维度（充分性和协调性）。随后，我们结合双元维度框架，基于3939家中国上市公司2011-2020年的相关数据，按照行业、省份、所有制、资本市场分类别、分阶段（“十二五”、“十三五”）统计、分析和描述企业共同富裕的双元维度水平和结构形态的演进历程。最后，我们还系统地提出了相关的管理建议和研究展望。
戴勇：提前期需求方差预测：对安全库存计算的影响
摘要：提前期需求预测是库存控制的基础。尽管已经有大量关于预测平均提前期需求的研究，但围绕预测提前期需求方差的研究较少，特别是在随机提前期的情况下。这代表了文献中的一个重要空白，因为安全库存计算明确依赖于提前期需求方差（或等效于无偏估计器的提前期需求预测误差方差）。我们通过探索在随机提前期下估计提前期需求预测误差方差的三种策略的可行性来弥合这一差距：（1）汇总提前期内预测误差的每周期方差，这是经典方法；（2） 考虑累计（提前期内）预测误差的方差；（3） 考虑到时间累计（在提前期长度上）需求导致的预测误差的方差。对于单指数平滑和最小均方误差预测方法，导出了一阶自回归移动平均ARMA（1,1）需求过程的分析结果。数值分析评估了需求自相关和提前期变化对每种策略准确性的影响，以及一种策略优于其他策略的条件。结果表明，教科书中提出的经典策略似乎是最不准确的策略，但具有高负需求自相关的情况除外。对库存控制性能的分析还表明，对于正自相关，经典策略通常导致更高的库存成本和更低的服务水平。
沈超海：需求预测算法、中国汽车企业社会责任履行与混合寡头垄断竞争
摘要：本研究在中国汽车市场需求存在不确定性、政府制定关税的框架下，通过建立了一个混合寡头垄断竞争的博弈模型，来探讨能改善需求预测的算法将如何影响社会福利、以及对企业自愿承担的企业社会责任的份额的影响。本研究发现，从整体上，算法可靠性的提高不改变预期的市场均衡价格、产量，但是提高预期的消费者剩余和企业利润。
第二场
时间：2022年10月19日（周三）13：00
腾讯会议：656-872-986
报告形式：线上
报告人、报告题目及报告概要：
朱松：评级策略对评级机构市场份额的影响研究
摘要：近年来我国的信用评级行业竞争愈发激烈，而2021年监管层“取消强制评级”的通知又使得评级机构的业务压力再次提升，为此评级机构需采取种种评级策略来维护市场地位。基于此背景，本文以2016年到2020年在我国信用债为样本，研究了评级结果策略、评级收费策略和主动评级策略对信用评级机构的市场份额的影响，结果发现偏高的评级结果策略、偏低的评级收费策略能帮助评级机构占据更大的市场份额，而主动评级策略对信用评级机构提升市场份额并无帮助，甚至会导致市场份额下降。此外，各评级策略之间也会相互作用，声誉机制也会对评级策略的效果产生影响。本文旨在呼吁行业内能建立起以声誉和评级质量为导向的评价体系，并通过法律法规、行业自律等手段指导合理定价，同时提高具有公益性质的主动评级的业务地位，帮助信用评级行业提升整体素质，尽快与国际水平接轨。
刘白璐：GDP growth incentives and pollution emissions: evidence from China
摘要: We investigate the relationship between gross domestic product (GDP) growth incentives at the government level and pollution emissions at the firm level. Using a sample of Chinese firms in heavily-polluting industries from 2011 to 2020, we find that firms in the provinces with lower GDP growth rates are more likely to emit SO2 than firms in the other provinces. We also find that the results are stronger for firms in provinces with strong government power and in provinces where the promotion incentive of a government official is high. The results are weaker for firms when environmental regulation is weak and in the post-campaign period. We further find these firms are more likely to reduce environmental investment while increasing coal consumption. Overall, this paper provides systematic evidence on how GDP growth incentives trigger pollution emissions at the firm level.
邓英雯：The Impact of International Trade on Auditor Choice: Evidence from China
摘要: We investigate the effect of companies’ international trade on their auditor choices. Using a sample of Chinese listed firms from 2001 to 2016, our results show that companies with more international trade (exports or imports) are more likely to choose Big 10 aud it firms, consistent with the signaling theory. To address the endogeneity concern, we perform a difference-in-differences (DiD) design, two-stage instrumental variable (IV) regressions, and the propensity score matching (PSM) method. Additional analyses show that the relation between international trade and auditor choice exists mainly in companies with low financial risk, supporting the signaling argument. Also, this relation is more pronounced (a) when companies’ overseas trading partners are from countries with a stronger legal environment and (b) when companies’ overseas trading partners are in countries with higher information transparency, but is weakened after the IFRS convergence in China. Overall, our findings verify the signaling role of large audit firms in the international market and enrich the literature in the field of the role of audit in cross-border supply chains.
第三场
时间：2022年10月19日（周三）13：00
地点： 腾讯会议：949-701-202
报告形式：线上
报告人、报告题目及报告概要：
楼雯：The diversity of canonical and ubiquitous progress in computer vision: A dynamic topic modeling approach
摘要：Research trends are the keys for researchers to decide their research agenda. However, only few works have tried to quantify how scholars follow the research trends. Here, we address this question by proposing a novel measurement for quantifying how a scientific entity (paper or researcher) follows the hot topics in a research field. Based on extended dynamic topic modeling, the degree of hotness tracing of papers and scholars is explored from three different perspectives: mainstream, short-term direction, long-term direction. By analyzing a large-scale dataset, containing all more than 270,000 papers and 45,000 authors in Computer Vision (CV), we found that the authors’ orientation is more in the established mainstream rather than based on incremental directions and make little difference in the choice of long-term or short-term direction. Moreover, we identified six groups of researchers in the CV community by clustering research behavior, who differed significantly in their patterns of orientation, topic selection, and impact. This study provides a new quantitative method for analyzing topic trend and scholars’ research interests, capturing the diversity of patterns of research behavior from a new point of view.
王琦萍：Does the Interplay between the Personality Traits of CEOs and CFOs Influence Corporate Mergers and Acquisitions Intensity? An Econometric Analysis with Machine Learning-based Constructs
摘要：Although the upper echelons theory posits that senior executives’ personal characteristics influence firm performance, very few studies have examined the impact of the interplay between CEO and CFO characteristics on corporate activities. To fill this research gap, we propose an econometric analysis model to examine the interplays between the personality traits of CEOs and CFOs and corporate mergers and acquisitions (M&A) intensity. In particular, our econometric analysis model is empowered by novel personality constructs extracted using a state-of-the-art machine learning-based personality detector that automatically mines CEO/CFO personality traits from firms’ earnings call transcripts. Based on historical M&A data of S&P 1500 firms, our econometric analysis reveals that the “openness” personality trait of CEOs is positively associated with corporate M&A intensity, while CEOs’ “consciousness” and “neuroticism” personality traits are negatively associated with corporate M&A intensity. Moreover, the impacts of CEOs’ “openness” and “consciousness” personality traits on corporate M&A intensity are more pronounced when CFOs have similar personality traits to those of CEOs.
贺国秀：Re-examining lexical and semantic attention: dual-view graph convolutions enhanced bert for academic paper rating
摘要：Automatically assessing academic papers has enormous potential to reduce peer-review burden and individual bias. Existing studies strive for building sophisticated deep neural networks to identify academic value based on comprehensive data, e.g. , the academic graph and the full paper. However, these data are not always readily accessible. And the content instead of other features outside the paper should matter in a fair assessment. Besides, though BERT models can maintain general semantics by pre-training on a large-scale corpus, they tend to be over-smoothing due to stacked self attention among unfiltered input tokens. Therefore, it is nontrivial to figure out the distinguishable value of an academic paper from the limited content. In this study, we propose a novel deep neural network, namely Dual-view Graph Convolutions Enhanced BERT ( DGC-BERT ), for academic paper acceptance estimation. We combine the title and abstract of the paper as input. Then, a pre-trained BERT model is employed to extract the paper's general representations. Apart from the hidden representations of the final layer, we highlight the first and last few layers as lexical and semantic views. In particular, we re-examine the dual-view filtered self-attention matrices via constructing two graphs, respectively. After that, two multi-hop Graph Convolutional Networks (GCNs) are separately employed to capture the pivotal and distant dependencies between tokens. Moreover, the dual-view representations are facilitated by each other with biaffine attention modules. And a re-weighting gate is proposed to further streamline the dual-view representations with the help of the original BERT representation. Finally, whether a submitted paper could be acceptable is predicted based on the original language model features cooperated with the dual-view dependencies. Extensive data analyses and the full paper based MHCNN studies provide insights into the task and structural functions. Comparison experiments on two benchmark datasets demonstrate that the proposed DGC-BERT significantly outperforms alternative approaches, especially the state-of-the-art model MHCNN and BERT variants. Additional analyses reveal the significance and explainability of the proposed modules in the DGC-BERT. Our codes and settings have been made publicly available on Github.
金武刚： 图书馆：公共文化服务均衡发展的引领者和促进者
摘要：推动城乡一体、全域均衡发展，实现人群全覆盖，是当前公共文化服务高质量发展的主攻方向。图书馆作为公共文化服务重要行业之一，在促进城乡均衡、人群均衡等方面率先示范，为现代公共文化服务体系建设探索了路径、创新了机制，成为我国公共文化服务均衡发展的引领者和促进者。面向未来高质量发展，图书馆可以在时间互通、空间互嵌、服务互融等方面继续统筹创新，促进公共文化服务向优质均衡方向迈进。
袁毅：元宇宙组织管理新模式
摘要：元宇宙组织管理新模式。摘要：元宇宙是个虚拟与现实交互的环境，企业等组织管理模式将发生重大的变化。近十年多以来，一些大公司致力于科层制向扁平化转型，但成效甚微。WEB3.0的到来，区块链技术等，为扁平化或去中心化管理模式提供了技术支持。 本报告主要是讨论新的技术背景和社会环境下的组织管理模式。
王仁武：音乐情感分析及应用
摘要：在商业服务情绪的研究中，文本情感分析已经得到了广泛深入的研究。近年来图片视频及音乐的情感分析，正在越来越多的得到业界和学界的重视。本研究报告将尝试对音乐情感分析的理论方法，数据集及应用作出研究梳理，以期抛砖引玉。
阮光册：自然语言处理之主题识别
摘要：文本主题识别处理的对象是典型的无结构数据，其核心问题是如何融入上下文知识，理解文本的语义信息。从空间向量模型到主题模型再到神经网络的深度学习模型，自然语言处理技术不断发展。汇报将围绕自然语言处理技术的发展过程，介绍文本主题识别方法的发展及演进方向。
蔚海燕：数据重用视角下的研究型图书馆数据服务
摘要：科学数据开放与共享是目前全球科技政策关注的重点，科学数据开放获取的目的在于重用数据。数据重用目前还是一个未被充分研究的问题，需要更多关注（Pasquetto，2017）。在科学数据管理与共享的运动中，出版商、科学数据中心、专业/行业数据仓储机构、高校以及研究型图书馆都积极制定科学数据管理政策、建立数据存储平台等。当全球的各类开放科学数据平台逐步建设起来后，用户面临的问题是：如何在海量的开放科学数据中发现最能匹配自己需求的数据。研究型图书馆作为学术资源的组织者和服务提供者，如何为科研用户提供科学数据服务，满足他们的数据重用需求，是现在研究型图书馆亟需解决