计算社会科学视角下的数据观

以大数据为标志的数据密集时代的到来，为社会科学实证研究提供了丰富的研究资料和全新的数据处理技术，计算社会科学迎来了一次重大的发展机遇。但当前大数据的概念混乱和概念泛化导致不少社会科学研究者对数据密集型知识发现以及计算社会科学的理论、方法出现认识偏差。通过对大数据的本质和特征的辨析，从计算社会科学视角厘清大数据的定义；并分析和阐释与大数据密不可分的以物联网、非关系型数据库、云计算、人工智能为代表的新一代信息技术体系，以及二者之间一体两面的关系。在此基础上，依据现实数据状态与数据处理技术水平，分析当前数据驱动型知识发现和理论数据双向驱动型知识发现的优势和局限，计算社会科学研究中的所面临的困难及突破方向。

Concept of Data from the Perspective of Computational Social Sciences

The advent of a data-intensive era marked by big data has providedabundant research materials and new data processing technologies for the empirical research in social sciences,and the computational social sciences have won a major opportunity for development.However,the current confusion and generalization of the concept of big data have led to many cognitive deviations in the data-intensive knowledge discovery and the theories and methods of social sciences.This paper clarifies the definition of big data from the perspective of social sciences by analyzing the nature and characteristics of big data,a new generation of information technology system represented by IoT,non-relational database,cloud computing and artificial intelligence which are closely related to big data,and the relationship between the two.On this basis,according to the status of real data and the level of data processing technology,it illustrates the advantages and limitations of the current data-driven knowledge discovery,the knowledge discovery of bi-direction-driven theoretical data,and the difficulties and breakthrough directions in computational social sciences research.