Font Size: a A A

Research On Knowledge Diffusion From The Perspective Of Heterogeneous Two-layer Academic Network

Posted on:2023-10-03Degree:DoctorType:Dissertation
Country:ChinaCandidate:L SongFull Text:PDF
GTID:1529306614990819Subject:Management decision-making theory and application
Abstract/Summary:PDF Full Text Request
Scientific research activities are the basis of knowledge innovation,and scholars’ academic behavior is an important way of knowledge diffusion.Researchers usually use academic data to analyze and explore the knowledge diffusion.The two most important mainstays in academic data are scholars and papers.Scholars are an important driving force to promote knowledge diffusion,and papers are the carrier of recording knowledge.As an intermediate link connecting knowledge creation and knowledge application,knowledge diffusion plays a vital role in knowledge inheritance and innovation.It can not only improve the efficiency of knowledge application in the whole society,but also promote the generation of knowledge innovation.Knowledge can be divided into explicit knowledge and tacit knowledge.Before the academic paper is published,knowledge exists in the minds of scholars in the form of tacit knowledge.When the paper is published,tacit knowledge is manifest and recorded in the academic paper in the form of explicit knowledge.The diffusion of knowledge includes both explicit knowledge diffusion based on papers,and tacit knowledge diffusion represented by the communication and cooperation among scholars.Therefore,knowledge diffusion shows the multidimensional nature of the process and the multiplicity of paths.How to use multi-layer network modeling to analyze the diffusion of explicit knowledge and tacit knowledge is a meaningful research work.Firstly,for explicit knowledge diffusion,this work explores the topic evolution of papers produced by explicit knowledge diffusion in the process of scholars’ cooperation,takes topics shown in academic papers as the quantification of explicit knowledge,and uses the topic evolution to characterize the diffusion process of explicit knowledge.From the perspective of network structure,a scholar-paper heterogeneous two-layer network model is constructed.The structure model is composed of the co-author network and the paper similarity network.The coupling relationship between layers is the process of knowledge manifestation by scholars.According to the structural evolution of the two-layer network,it is found that old scholars and old papers are very important for the cohesion and development of discipline knowledge.In order to further explore the knowledge topics displayed in papers,through the community detection of the paper similarity network,the paper community gathered from the same topic is obtained,so as to build the scholar-topic two-layer network model,realize the coupling representation between scholars and their research topics,and supply the vocabulary association network and topic evolution network to assist in obtaining the specific content of the topic.From the perspective of state,the state of scholars(temporary,growth and reputation)and the obtained topic state(new,divided,self-expanded and dormant)are defined by using the length of scholars’ career and the evolution type of topics.According to the coupling evolution process of the two-layer network,it is found that scholars in all states generally have low contributions to new topics,self-expanded topics and dormant topics,and scholars pay more attention to divided topics,among which reputation scholars have the greatest contribution to divided topics.Therefore,reputation scholars have a more significant impact on promoting the explicit knowledge diffusion.Secondly,this work further studies the tacit knowledge diffusion,which carries more knowledge value than explicit knowledge.Aiming at the problems of many subjective variables and large fluctuation of results in the dynamic method most commonly used to evaluate the tacit knowledge diffusion,an algebra matrix evaluation method is proposed,which uses the algebra matrix to integrate the social network structure with the state of tacit knowledge owners and explicit knowledge carriers.According to the local micro diffusion process of tacit knowledge,the self-consistent equation is constructed to evaluate the scale and threshold of tacit knowledge diffusion.A Monte Carlo simulation experiment is designed combined with the real situation.By comparing the simulation results in artificial networks with different degree distribution and real social networks,the effectiveness of the algebra matrix evaluation method is verified.In the process of analyzing the real data,a scholar-paper two-layer network model composed of coauthor network and paper citation network is constructed.The states of tacit knowledge owner and explicit knowledge carrier are defined by using the career length of scholars and the number of citations of papers,respectively.Through the study of real data,it is found that in the process of the continuous development of cooperative diversification,scholars are more willing to exchange their tacit knowledge,which is very beneficial to the development of science.Further research shows that the scale of cooperation is an important factor to promote the tacit knowledge diffusion.The low scale of cooperation seriously hinders the tacit knowledge diffusion and limits the development of scientific research.In addition,by comparing with the dynamic evaluation method,it is found that the influence of state information on tacit knowledge diffusion is significant.In the largest connected component,the evaluation deviation caused by ignoring state information is13.44%,and in all co-author networks,the deviation increases to 45.07%.Finally,focusing on the micro research on the identification of key scholars,this work analyzes the impact of key scholars on knowledge diffusion.In order to overcome the one sidedness of single index evaluation and the redundancy of multi index evaluation to realize the comprehensive,objective and efficient evaluation for scholars,this work proposes a structural index reduction algorithm(SIRA),which completes the comprehensive evaluation of network nodes through four processes: structural index selection,index equivalence class construction,representative index extraction and node comprehensive evaluation.The algorithm belongs to multi index evaluation algorithm,which does not consider the calculation of structural index,and the time complexity of the algorithm itself is only.From the perspective of structural robustness,including network connectivity and network efficiency,it is found that SIRA algorithm has good recognition effect on key nodes.From the perspective of the characteristics of key scholars,including growth,extensive cooperation and sparse cooperation among key scholars,it is found that SIRA algorithm can accurately identify real key scholars.Key scholars play a leading role in explicit knowledge diffusion due to their maintenance of cooperative relations and high support for various topics.In addition,key scholars have made great contributions to the division of topics and the generation of new topics,which makes them promote the explicit knowledge diffusion.Accordingly,comparing the influence of key scholars and non-key scholars on the tacit knowledge diffusion,it is found that key scholars play an important role in promoting the diffusion of tacit knowledge and are an indispensable part of cooperation and communication.According to the above research contents and main research conclusions,the innovative contributions of this work are summarized as follows:(1)By studying the complex structure formed by heterogeneous objects,this work puts forward a new research perspective from the structural level of heterogeneous two-layer network,which opens up new ideas for analyzing the mechanism of knowledge diffusion;(2)According to the multi-dimensional information of knowledge contained in heterogeneous two-layer network,an algebra matrix evaluation method which can fuse multi-dimensional information is creatively proposed,and this new method is used to evaluate knowledge diffusion;(3)According to the theory of structural embeddedness,an index reduction algorithm is creatively proposed to avoid the redundancy of multi index evaluation,and this new algorithm is used to accurately identify the key scholars who play a leading and promoting role in knowledge diffusion;(4)Using the new perspective,the new method and the new algorithm,this work explores the influence mechanism of knowledge owners and knowledge carriers on knowledge diffusion from the levels of structural evolution and state evolution,so the research conclusion has strong theoretical innovation.The research topic of this work comes from the National Natural Science Foundation Project“Multi-layer social network theory with application based on heterogeneous information fusion”(No.72171136).From the perspective of heterogeneous two-layer network,this paper constructs a two-layer academic network model to realize the integration of multi-dimensional heterogeneous information of knowledge.By studying the structure and evolution characteristics of heterogeneous two-layer academic network,the diffusion mechanism of explicit knowledge and tacit knowledge,and the micro impact of key nodes in the network on knowledge diffusion,this work greatly expands the perspective of knowledge management research and enriches the tools of knowledge diffusion evaluation.In the research process of knowledge diffusion mechanism,through selecting the new perspective,proposing the new method,designing the new algorithm and other innovative work,we explore the research conclusions with strong theoretical innovation,which provides a reference for the management optimization and decision-making of scientific research work.
Keywords/Search Tags:Knowledge Diffusion, Two-layer Academic Network, Topic Evolution, Algebra Matrix Evaluation, Key Scholar Identification
PDF Full Text Request
Related items