The Revelation Model Building Of Knowledge Structure And Its Empirical Research

Purposes:From the perspective of Medical Informatics,this study aims to integrates the systematic knowledge structure research paradigm,constructs systematic revealation model of knowledge structure,forms standardized revealation process and knowledge map method of knowledge structure.Through the construction of revealation model,explore the model of each module coordination mechanism and the overall operation process.At the same time,an empirical study of type 2 diabetes mellitus is carried out to demonstrate the scientificalness,rationality and operability of the model,providing decision support for disease prevention,diagnosis,control and treatment,as well as enriching clinical medical evidence.Methods:First of all,this study systematically reviewed the research status of knowledge structure at home and abroad through literature review,and evaluated and explained the theories,methods and technologies related to knowledge structure.Secondly,the knowledge structure is innovatively graded,which is divided into low level,middle level and high level,making the research of knowledge structure more targeted and hierarchical.At the same time,summarizing the two research paradigms of knowledge structure revealed by "three metrics"(Bibliometrics,Scientometrics and Informatics):the "Representation method" starting from the external characteristics of discipline and the "Connotation method" starting from the content characteristics of disciplines.Finally,constructing the knowledge structure revelation model – the "233-1 model",combing out the general running process of revealing the model : data collection,data processing and knowledge structure interpretation.Finally,based on revealation model of knowledge structure and running process of model,using type 2 diabetes as the revealing body of knowledge structure to reveal the dynamic evolution and knowledge discovery process of the disease.Based on knowledge map,using social network analysis,time-keyword co-occurrence analysis,time slice,topic classification,words frequency statistics,the dynamic evolution of type 2 diabetes mellitus was studied.As to knowledge discovery in type 2 diabetes mellitus' s high-level knowledge structure,based on knowledge map,using social network analysis and topic keyword co-occurrence analysis,research on two aspects : theme independence and theme relevance.In the study of theme independence,Theme Relative Independence Algorithm was proposed to find the difference between diabetic nephropathy and diabetic retinopathy.In the study of theme relevance,Theme Association Intensity Algorithm is proposed to explore the differences in the similarity of diabetic nephropathy and diabetic retinopathy.Results:? Build a revelation model,named "233-1 model",which on the basis of "three metrics " module,takes social network analysis module as the support,and discipline knowledge structure module as guide and knowledge map module as the output;? Explor the running process of the revealation model,including three steps:data acquisition,data processing and interpretation of knowledge structure,and gives a specific description to different levels of knowledge structure;? Reveal the overall heat evolution in type 2 diabetes from 2002 to 2017 in the 15 year time window,and extracting three development periods:stationary phase,surge period and decreasing phase.And it interprets the main hotspots in three periods from two dimensions;? Extract four types of evolution themes of type 2 diabetes:continuous type,dilated type,dissipated type and jumping type.And reviewed and verified the Theme Evolution types,including drug research progress,treatment methods,target and pathogenesis;? Explor the knowledge discovery mode of type 2 diabetes:based Theme Relative Independence Algorithm and Theme Association Intensity Algorithm.Taking diabetic nephropathy and diabetic retinopathy as an example,excavat the difference and similarity of two kinds of microangiopathy.Conclusion:? In accordance with the principles and ideas of the "233-1 model" and the model running process,the empirical research proves that it is scientific,reasonable and effective.The revelation model and running process consummate the reveal process of knowledge structure,concreting the implicit and abstract knowledge structure to form a more rigorous method standard and theoretical paradigm,provide theoretical support for empirical research,and also provide methodological reference and basis for the personnel engaged in related research;? The evolution trend of a discipline is explored from macro thermal evolution,can intuitively and systematically grasping the historical track and development path of the discipline.Micro thermal evolution based on time slice and word frequency statistics can further explore the main research hotspots,secondary research hotspots and the trend of overall hot spots in different periods;? By classifying,visualizing and interpreting the theme evolution trend,can clearly reveal the theme evolution state and evolution path of type 2 diabetes in the whole time series.In the prior case,through the heat evolution and theme evolution method,it can trace the trend,industry dynamics and development of medical field in advance,and improve the treatment methods,track the international advanced thinking and technology,and match the industry standard;? The Theme Relative Independence Algorithm and Theme Association Intensity Algorithm are effective and feasible for theme independence research and theme relevance research in knowledge discovery.The theme independent research can be found in two kinds of similar disease their specificity,operation and treatment indications,professional treatment guidance,drug indications,common risk treatment mechanism.The theme association research can be used to explore the common mechanism of disease treatment,common risk factors and potential treatment methods.