Design And Implementation Of Knowledge Engine For Chronic Disease Rehabilitation Of Elderly People

In recent years,the number of chronically ill people in China has risen steadily,which poses a great threat to the health of the people.The elderly population,as a key target for chronic disease rehabilitation,needs to use a large number of offline rehabilitation resources such as rehabilitation institutions and rehabilitation experts.On the one hand,offline chronic disease rehabilitation resources are unevenly distributed,and the degree of informatization is low.It is difficult to obtain chronic disease rehabilitation services,which leads to a slow development of elderly chronic disease rehabilitation management.On the other hand,pension-oriented auxiliary medical service robots are the hotspot of current research and application,but at present there is a lack of basic technology and methods to support such service robots.In order to improve the utilization and service quality of chronic disease rehabilitation service resources,provide basic service capabilities for downstream chronic disease rehabilitation knowledge question-answer system,and chronic disease rehabilitation service recommendations,etc,this paper builds a knowledge engine for elderly chronic disease rehabilitation.From three perspectives of the construction of the chronic disease rehabilitation knowledge graph,the identification of chronic disease rehabilitation user intentions and the recommendation of the chronic disease rehabilitation plan,this paper studies the corresponding theoretical methods.The specific contents are as follows:Firstly,the data support of the chronic disease rehabilitation knowledge engine is the chronic disease rehabilitation knowledge graph.This paper studies the knowledge graph construction method in the field of chronic disease rehabilitation,determines the topdown knowledge graph construction process,and gives the chronic disease rehabilitation ontology definition of the domain,expounding the method of extracting chronic disease rehabilitation knowledge based on the wrapper,using the native graph database Neo4 j to store knowledge data.After analyzing the data structure of the knowledge graph,this paper uses translation model TransD to obtain the distributed representation of the knowledge graph.Secondly,the intent recognition is the interface between the knowledge engine and the user,and is the basic service component of the dialogue robot and question answering system.This paper proposes a concept co-occurrence enhanced ERNIE model using both semantic features and concept co-occurrence features for intent recognition.The experimental results show that the model has a greater performance improvement than the existing intent recognition model.The model can effectively solve the problem of recognition of multiple intentions and implicit intentions.Finally,this paper designs the architecture and functional structure of the chronic disease rehabilitation knowledge engine system,and designs the main functional modules in detail based on the research methods above.The system verifies the feasibility of the methods we proposed.