A Research Of Scientific Field Of Study Based On Complex Network

Driven by big data,Scientometrics computes the academic society quantitatively,and understands the patterns and laws of science development by complex network modeling using massive data.From a perspective of complex network,this paper studies patterns of scientific fields of study development,and structural and evolutive properties of field of study network through analyzing the journal data in Microsoft Academic Graph database and American Physical Society offical dataset.The results are as follows:First,it is demonstrated that scientific fields of study which a paper involves in tend to rise on average as time goes by.This phenomenon also matches to the growing specialization and cooperation in science.Besides,the number of papers among fields obeys power law distribution or truncated power law distribution,which indicates that scientists have a preference for hot research fields,namely “following the popularity effect”.Based on hypotheses above,we propose a paper publishing process model,whose theoretical and simulative results explain the power law partly.Also,this paper constructs a scientific field of study network based on their relationships.By studying architecture and evolution of it,we find a scale-free character,which means its degree and coreness both follow tailed power law distribution.Furthermore,we found that some dense core which composed of several nodes with large degree connect with each other.Apart from that,some key indicators of the network,such as degree distribution,coreness distribution,average cluster coefficient,average shortest path length and degree assortativity coefficient are relatively stable over time,which show a relatively stable scientific field of study network.Finally,according to citation among fields,this paper constructs an inter-citation network and proposes a paper growth model based on citation flow function.The model can predict the number of papers of fields in a few years,using inter-citations as well as publications data.Hence,it help determine the future development of scientific fields.All the findings in this paper demonstrate that both development of fields of study and relationship among them have their inherent disciplines.These disciplines can help people to better predict the development of fields of study and assist scientific policy makers to formulate policies that can support and promote fields of study development more reasonably.