中文摘要：

随着大数据在各领域的广泛应用，“数据密集型科学”成为新的研究范式，也为投资者群体行为的度量提供了新的数据和方法支持。本文以网络大数据的来源和投资者行为为分类标准，从投资者关注、信息需求、情绪和公司盈利预测四个方面对基于网络大数据的投资者群体行为及其与股票市场的关系的相关研究进行了梳理、总结和述评，发现网络大数据对资本市场具有重要的信息价值，网络大数据中所反映的个体偏好和预期及其传播和演化为市场信息反馈和投资者行为的研究提供了很好的实证数据。同时，计算机科学与金融学的交叉研究在其中具有重要的作用。本文在现有研究的梳理上提出基于网络大数据的投资者观点和行为的演变、交互，资产配置策略，市场参与者的情绪及其影响机制，资本市场的监管以及结合公司金融、外汇市场等其他金融市场的研究可能是未来的研究发展方向。

英文摘要：

With the widespread application of Big Data, ＂data-intensive science＂ has become the new research paradigm, which has provided new database and methodology for collective behavior measure of in- vestors. By using the source of Internet Big Data and investors＇ behavior as the classification criteria, this pa- per reviews the studies on investors＇ attention, investors＇ information demand, investors＇ sentiment, and investors＇ earnings forecasts based on Internet Big Data. According to the review, we find that Internet Big Da- ta has important information value on capital markets. The individual preferences and expectations reflected in the Internet Big Data provide good datasets for empirical research on market information efficiency. Besides, the interdisciplinary research of computer science and finance has played an important role in finance stud- ies. This paper proposes that the evolution and interaction of investors＇ preference and behavior, portfolio man- agement sentiment of market participants and market supervision as well as applications on other markets and corporate finance based on Big Data are the possible trend of future research.