Research On Knowledge Mapping Construction For Diagnostic Diagnosis Of Short Stories

Due to the coming are of big data,knowledge map has been an important operation tool for knowledge management and application since it can be used to express semantics,retrieve and deduction in a wide variety of searching engines.The knowledge map modeling is comprised of a series of procedures such as named entity extraction,knowledge fusion,entity disambiguation,relationship extraction,knowledge representation and knowledge reasoning.Compared with other semantic network,it can be more efficient and standard when used to make a big-scale storage of professional knowledge.Many large and relatively complete bilingual knowledge map has been made in some general fields by far,but the technology is still immature.Take medical diagnosis as an example,although numerous China's traditional medical knowledge map has been built,it was far behind in auxiliary diagnosis using,which greatly hindered the development of medical intelligence assisted diagnosis.In this paper,according to diagnostic criteria for short stature,a medical knowledge map on its auxiliary diagnosis was constructed,including the following 3 aspects :1.The diagnostic cores of the short stature was analyzed according to the diagnosis criteria and related data in the diagnosis manual and the data involved are detailed,as well as the short stature diagnosis process involve.2.Specifically analysis the key diagnostic steps and related knowledge involved in the diagnosis of short stature,and use CRF and jieba classifying and comparing them to compare and select the tools with higher accuracy of word segmentation,then got the short stature diagnosis' s main features to determine the ontology.3.Perform knowledge representation and deduction,and make an analysis and introduction on the key techniques used in detail.Use Protégé ontology editor to construct the entities and their attributes in the short stature knowledge map and Protégé to construct thecompleted knowledge graphical display.Based on the formation of the knowledge map,generate the OWL file through the ontology construction model,and then instantiate the data with OWL to generate the RDF file.Then,use the SWRL language to write inference rules and use the inference engine that comes with Protégé for deduction.Finally,the quality assessment and improvement of the completed knowledge map is carried out.