Research On Academic Promotion Of Drugs Based On Text Mining And Social Network

Medicine,as a special commodity,has its side effect.So,Drugs require strict control of the state.The most effective way to promote domestic drugs is academic promotion,corresponding to foreign clinical trials.Usually researchers can carry out clinical trials of drugs,speed medical knowledge to achieve the promotion of drugs.Because of the complexity and variability of medical terms,such as medicine,prescription,medicine,disease,syndrome and disease for clinical research literatures,the terminology extraction in medical field has become a difficult problem.In this paper,on the basis of Analysis of Medical Knowledge,we combine the user-defined dictionary and construct a Medical Term Extraction Model Based on Conditional Random Field.The medical terminology extraction model achieve the extraction of medical entities such as diseases,drugs,prescriptions,symptoms,treatment and so on.With the literature of clinical research as the object,we can realize the analysis of the promotion of medicine from multiple angles.First,this paper uses the method of combining text mining with multivariate statistical analysis.On the one hand,We extract terms such as diseases and medicines from the literature and analyze the evolution of disease and the development of drugs with time.Then,we detect the feedback of academic promotion corresponding to the population of the diseases comparing with the disease spectrum produced according to the "China Health and Family Planning Statistical Yearbook".On the other hand,through the analysis of the content of the literature,we find the language rules of medical terms,in addition we can further find the law of potential development of medicine.Secondly,based on the social network analysis method,this paper studies the potential cooperation network structure of drug researchers.Because the field of medicine is a highly specialized field,cooperation between researchers can be predicted reasonably although there is little cooperation in clinical research in the field of medicine in China.The network structure in this paper presents the characteristics of small world network and obvious community structure.At present,the academic research on drugs is mainly some descriptive analysis and possibility analysis.The academic promotion of drugs depends on the clinical experiment of clinical research scholars and knowledge marketing from an academic point of view.In this paper,we study the situation and trend of the academic promotion of drugs from the point of view of text mining and social network.On the one hand,it is helpful for us to know how to promote the academic development of drugs from the data and deepen our understanding of the behavior and relationship of drugs,diseases,and the promotion of medicines level.On the other hand,we can predict and find potential collaboration networks among drug researchers to provide theoretical help for future collaboration.Drugs are a special commodity,"is the drug three drug" characteristics make the drug needs the country's strict control.The most effective way to promote domestic drugs is academic promotion,corresponding to foreign clinical trials,can be through the clinical trials of drugs,medical knowledge will be spread to achieve the promotion of drugs.Among them,the medical field of clinical research literature,medicine,prescriptions,herbs,disease,card,disease and other medical terms of the complexity of the change in the field of medicine makes the terminology extraction has become a difficult point.In this paper,based on the analysis of medical knowledge,combined with user-defined dictionary,given the conditional random field based on the medical term extraction model,the medical term extraction model to achieve the disease,drugs,prescriptions,symptoms,therapy and other medical entities Extract.To the clinical research literature as the object,we can achieve from a number of angles to achieve the promotion of pharmaceutical academic analysis.First,this paper uses the method of combining text mining with multivariate statistical analysis.On the one hand,the terminology such as disease and medicine is extracted from the literature and the time series,the evolution of the disease with time and the development of the drug Regularity,and in accordance with the "China Health and Family Planning Statistical Yearbook" produced a comparative analysis of the disease spectrum,to explore the academic promotion of population-related changes in disease feedback.On the other hand,through the analysis of the content of the literature,to find the language of the language of language,in addition we can further find the potential development of medicine law.Secondly,this paper studies the potential cooperative network structure of drug researchers based on the social network analysis method.The cooperation between the clinical research in the medical field of our country is particularly small,but the medical field is a strong field in the field of cooperation.Very reasonable prediction,so this paper established a potential cooperative network of drug researchers,this network structure showing the characteristics of a small world network,and there are obvious community structure.At present,the academic research on drugs is mainly some descriptive analysis and possibility analysis.The academic promotion of drugs depends on the clinical experiment of clinical research scholars,from the perspective of academic knowledge marketing.This article from the perspective of text mining and social network of academic promotion of the situation,the trend of research,on the one hand,is conducive to us from the data level to understand the academic development of drugs how to deepen our drugs,diseases,drugs academic promotion Behavior and understanding of relationship.On the other hand,we can predict and find a potential network of cooperation between drug researchers to provide theoretical help for future collaboration of researchers.The main contents of this paper are as follows:(1)Based on the characteristics of clinical research literature and combined with complex medical field knowledge,a conditional random field model for identifying medical terminology of clinical research literature is presented,and the above respiratory tract infection disease is taken as an example to realize entity recognition.This model combines the two methods based on dictionary and statistics,and has the advantages of both.Recognition results are superior,and have a certain degree of expansion,to a certain extent,can be extended to the identification of other diseases.The model is part of the machine learning,on the one hand through the model continues to increase the amount of medical terms dictionary,on the other hand the model for the subsequent academic promotion behavior analysis and social network analysis provides a data processing method,is the basis of the latter two parts.(2)Combining the methods of text mining and multivariate statistical analysis,the behavioral analysis of clinical research literature was carried out.The dimensions of the literature,the dimensions of the research institutions,the dimensions of the drugs,and the dimensions of the disease were analyzed in succession,and the data were visualized.The clinical research situation and the disease spectrum were compared and analyzed;and the contents of the literature were analyzed,and the evolution relationship between proprietary Chinese medicine and traditional Chinese medicine prescription was excavated.(3)The potential cooperative network of drug researcher in medical field was established,and the structural characteristics of cooperative network were analyzed.This network presents a small world network,but also belongs to the scaleless network,and can be divided into multiple communities.This paper also further analyzes the network structure of the cooperative network and finds the core figure and analyzes it.