Construction Of Knowledge Base For Health Management Program And Research On Intelligent Generation Of Personalized Programs

At present,the prevalence and mortality of cardiovascular diseases in China are still on the rise,some authoritative reports show that the number of cardiovascular patients nationwide exceeds 290 million,2 out of every 5deaths die of cardiovascular disease.Therefore,cardiovascular disease has become the number one “killer” of people's health in china,and the prevention and control work is necessary and no time to delay.Social development and many medical practices show that intelligent health management based on preventive maintenance,real-time monitoring and personalization is not only a common health solution,but also an effective strategy to deal with chronic diseases such as cardiovascular diseases.The rapid development and widespread popularity of the Internet have provided opportunities for knowledge sharing and also provided a guarantee for the realization of personalized health management.With the great progress of Internet technology,the knowledge of network-based health management has also exploded.However,because these massive and free sources of health knowledge are numerous,large,uneven in accuracy,and vary in expression,it is difficult for people to find the high-quality health management knowledge that is truly suitable for themselves.At present,there are many Internet-based public health management systems,but most of the health management knowledge released by these system platforms is qualitative and generalized,and is seriously insufficient in quantification and pertinence,and there is also a lack of clear implementation steps and processes.As a result,users can not effectively apply such knowledge to improve the health level for themselves.For the above problems,this paper proposes a domain ontology knowledge base model to standardize and represent the knowledge of health management programs in the field of cardiovascular diseases.Then,according to the designed domain ontology library,this paper combines individual health characteristics,environment characteristics and other individual characteristics to further study the intelligent generation algorithm of personalized health management programs.Finally,according to the qualitative personalized exercise and diet programs,the quantitative models of personalized exercise and diet programs are studied and proposed.This paper also invited domain experts to evaluate the personalized health management programs generated based on patient cases,and the evaluation results proved the effectiveness and scientificity of intelligent generation and quantification of personalized programs.The main innovations of this paper are as follows:First,this paper summarizes the key concepts related to health management in the field of cardiovascular disease,and proposes a domain ontology knowledge base model of health management programs,which makes up for the lack of domestic research in this field and helps to filter,extract,logicalize and structurize multi-source heterogeneous health management knowledge in this field.In addition,considering the various risk factors of cardiovascular diseases,the previous ontologies,which were limited to the field of disease and medication,could not well meet the personalized health management needs of the public.Therefore,this paper constructs the health management program ontology and health management program implementation ontology,as well as other key ontologies,such as food,recipe,exercise and other basic ontologies,and also contains many ontologies and attributes related to individual health characteristics.The domain ontology library constructed in this paper is a set of knowledge base model on the cardiovascular disease field,which completely aims at the personalized health management for the public.At the same time,it is also a basic research,which can be widely applied in scenarios related to health management in the future.Second,this paper applies fuzzy Petri networks to the field of health management,and uses the parallel inference method based on matrix operations to intelligently generate the health management programs,which solves the efficiency problem of large-scale knowledge inference.The proposed algorithm fully considers the individual health characteristics as well as the related natural environment and social environment characteristics,can mine some health characteristic information or other characteristic information ignored or missed by users.The algorithm can automatically identify and deal with the contradiction or conflict rules in the process of knowledge inference,and ensure the validity and scientificity of the program recommendation results.For the diet program,the recommendation algorithm further selects and sorts the diet set based on the information of individual income level,diet preference,diet taboo,etc.In addition,this paper classifies and recommends the diet programs according to the nine diet categories of Chinese Balanced Diet Pagoda.The algorithm not only emphasizes the individuation of diet,but also takes into account the diversity and balance of diet types.Third,a qualitative exercise program without definit exercise time can neither support the implementation of personalized health management,nor guarantee the effectiveness and scientificity of exercise.Based on the proposed qualitative exercise program,following the principle that the difference between the total daily energy consumption and the daily dietary energy intake is closest to the daily recommended net energy consumption,this paper constructs a nonlinear programming model of the personalized exercise program,and solves the model to obtain the quantified exercise time.This exercise time is also an important input parameter for the follow-up diet quantification,thus ensuring the close relationship between exercise quantification and diet quantification.Fourth,a qualitative diet program without definit dietary intake will greatly reduce the operability and individual compliance of the personalized health management programs,and cannot ensure the scientificity of daily dietary energy intake.Based on the proposed qualitative diet program and the quantitative exercise time,this paper constructs a goal programming model of personalized diet programs by following the principle of balancing dietary energy intake with the total daily energy consumption calculated according to the target BMI,taking the recommended intake of nine diet categories of Chinese Balanced Diet Pagoda as the model constraints.Exercise quantification and diet quantification are interdependent,and the model is more scientific and reasonable.Based on ontology theory,knowledge management theory and optimization theory,combined with the research results of sports medicine and nutrition,this paper proposes a domain ontology knowledge base model of health management programs.Then,a series of knowledge reasoning methods and algorithms are designed,which can intelligently generate health management programs that meet the personalized needs of users.At the same time,for the two most common health intervention programs such as exercise and diet,the conversion from qualitative programs to quantitative models is realized,which greatly improves the operability,effectiveness and accuracy of health management programs.