A tool kit of computational social sciences

By WANG XIUXIAO and LING CHAO / 04-11-2024 / Chinese Social Sciences Today

An Introduction to Computational Social Sciences

Amid the ever-changing and rising tide of digitalization, the traditional social science research paradigm is encountering increasing challenges in adapting to the new economic and social patterns represented by big data, cloud computing, and artificial intelligence. At this historical juncture, marked by hesitation and confusion, An Introduction to Computational Social Sciences, under the chief editorship of Lyu Peng, a research fellow from the Institute of Sociology at the Chinese Academy of Social Sciences, and Fan Xiaoguang, a research fellow from the School of Public Affairs at Zhejiang University, emerges with the pioneering vision of establishing “the first artificial intelligence class for social scientists.”

The book aims to introduce the research methods of computational social sciences to students and practitioners of humanities and social sciences, while also presenting the main topics of interest to those in computational science disciplines. To the first end, the book adopts a user-friendly approach for novices, or non-technical students who may be intimidated by advanced mathematical principles, formula derivation, and code writing. It deliberately omits certain technical details, starting from basics such as guiding readers through the installation of Python software. 

The primary target audience is undoubtedly more focused on the first group, namely students and professionals in the humanities and social science fields. The first chapter systematically reflects on the five methodological dilemmas facing empirical social sciences, namely macro and micro explanation, data and theory-driven development, correlation and causation, homogeneity and heterogeneity, and reliability and validity. As some scholars take research findings that are common sense as important innovations, the book tends to arouse the ambition of humanities and social science practitioners to break free from traditional research constraints with the help of computational social sciences and data science.

To ensure social scientists do not feel completely lost, the book not only provides thorough theoretical groundwork, but also introduces several “traditional” methods relatively familiar to them, such as regression analysis, cluster analysis, and social network analysis. However, the highlight of the book lies in the introduction of the latest approaches to computational science, such as machine learning, neural network analysis, natural language processing, computer vision, and agent-based modeling.

The book serves as a versatile beginner’s toolbox to unlock the research door of digital society. It encompasses novice-friendly yet well-established methods to get started, while not omitting the more advanced cutting-edge technologies, covering the major methods and latest advancements in computational social sciences. Most chapters are supplemented with case studies in the humanities and social sciences, Python code with annotations, and data sets and codes, which can be downloaded for free. 

Wang Xiuxiao and Ling Chao are from the School of Sociology and Psychology at Central University of Finance and Economics.