Philosophical Research In Computational Social Sciences

In February 2009,an article titled “Computational Social Science” was published in Science,marking the official birth of a new filed,computational social science.Computational social science has unique advantages over traditional social science.The former adopts methods such as computational modeling and big data to study human behavior and social interaction.Being able to study and solve problems that traditional social research methods cannot deal with,computational social science has shown enormous potential for growth and achieved fruitful results.However,as a discipline that is still in its early stages of development,computational social science is mainly treated as a research method in existing research which pays attention to its use as a technical approach to studying specific problems while research into its disciplinary orientation,scientific value and philosophical foundation is still insufficient.This dissertation studies computational social science as a whole,sorts out the background of and the causes for the birth of computational social science,and identifies related ontological,epistemological and methodological issues.On this basis,the dissertation clarifies the disciplinary orientation and advantages of computational social science,and reveals its scientific value as well as its limitations.This dissertation is divided into seven chapters.The introduction explains the significance of this study,defines the concept of computational social science from three aspects,demonstrates the necessity of philosophical reflection on computational social science,and provides a literature review.The first chapter expounds on the characteristics of social science and the limitations of traditional social science research methods.Traditionally,there has been many different and even opposing paradigms in the field of social science,but there has also been a lack of effective methods in terms of precise control of objects and data acquisition.The second chapter systematically sorts out the background,development process and disciplinary characteristics of computational social science,and,at the same time,makes a distinction and analysis of its ontological promise.Chapter Three focuses on analyzing computational social science research methods and their epistemological implications and discussing issues such as the relationship between methods,theories and data,the reasoning mode and objectivity in computational social science.The fourth chapter concentrates on the cognitive function of computational social science.It mainly expounds the characteristics,types and mechanisms of explanation and prediction in computational social science.Chapter Five makes a further reflection on computational social science from the perspectives of philosophy of science and philosophy of social science.Besides,this chapter clarifies the advantages of computational social science and its disciplinary value in the field of social science.The sixth chapter explains both the internal and external problems and challenges which computational social science is faced with.The main idea and innovations of the dissertation are as below.First,it expounds the origin and ontological presuppositions of computational social science.Driven by technology and data,people are required to understand a complex society,thus giving rise to the birth of computational social science as a discipline.Computational social science regards human society as an intricate information-computing network.With the ontological promise of information-computation,it studies social phenomena through information-computing ideas and data technology,which shows its distinctive interdisciplinary nature.Second,agent-based modeling and simulation are described.The dissertation demonstrates that by this means,the pattern of how agents produce collective behavior through obeying simple rules can be discovered,which is an ideal approach to studying how microscopic human behaviors form macroscopic social phenomena.The introduction of big data technology provides a strong data support and empirical basis for this approach,which can considerably expand the depth and breadth of social research and effectively promote the development of social theory.Third,the dissertation clarifies and analyzes the issue of objectivity in computational social science.It argues that this objectivity cannot be simply reduced to algorithmic objectivity or intersubjectivity.Due to the distinctiveness of social research,objectivity cannot be completely separated from subjectivity.Computational social science is a practical activity that integrates society and technology.Therefore,it is necessary to interpret the objectivity of computational social science from a practical perspective.Fourth,this dissertation points out that the explanation in computational social science contains abductive reasoning,which has unique advantages in studying complex and open social issues.It also analyzes and demonstrates that the explanation in computational social science belongs to generative explanation and that the latter essentially belongs to mechanism explanation.Fifth,computational social science has unique methodological value.Within the framework of interpretivism,it adopts the methods advocated by positivism to explain and understand social phenomena,and plays the role of changing and transforming society which is advocated by criticism.Using formal analysis tools to study social interaction,computational social science has the advantages of “unresponsiveness” and convenience in the aspect of data acquisition,and is able to resolve the debate between individualism and holism to a certain extent.Sixth,computational social science is also faced with some problems and challenges,such as implicit bias and algorithm interference that is easy to be overlooked in data acquisition,inversion of research purpose and research method,separation of data and conceptual theory,etc.As a result,there is still a need for experts and scholars in data science and social science to promote closer cooperation and deepen mutual understanding.In conclusion,this dissertation systematically and comprehensively discusses the background of computational social science,its research methods,disciplinary nature and status,and the philosophical presuppositions involved in this discipline.This study will contribute to further clarifying the scientific value of computational social science,facilitating the integration between computational social science and traditional social science,and promoting the development of social research as well as the construction of a smart society.