论社会学理论导引的大数据研究——大数据、理论与预测模型的三角对话

On Big Data Research Guided by Social Theory——The Three-way Interplay of Social Theory,Data Mining and Predictive Models

计算社会科学把社会科学理论以及研究方法与大数据分析熔为一炉,一方面为大数据分析开启了很多新议题,理论指导下的定性、定量调查也可以为数据挖掘的结果提供校准的扎根真相;另一方面,在大数据挖掘的结果中可以找到建构理论的线索,提供验证理论的资料,进而指导预测模型的建构,推论并解释更多的现象。本文以中国风险投资产业网络数据为例,展示了数据挖掘、社会学理论与预测模型间的三角对话,进一步呈现了以理论导引的大数据分析的方法论。

Due to advances in big data analysis,computational social science is now firmly in the spotlight of social science research.This rapidly emerging field integrates social theory and data mining,giving researchers many new tools to use and research topics to explore.In the research process,social theories provide an overarching framework to guide researchers use of qualitative and quantitative surveys.With these aids researchers collect ground truth to test the results of data mining.In turn,these results provide evidence for researchers to build new theories.Big data can also be used to test new theories,which helps construct predictive models and infer new“facts”.The three-way interplay of social theory,data mining and predictive models is the background for this paper s examination of Chinese venture capital firms industrial network data.