Research On The Construction Method Of Civil Aviation Uncivilized Passenger Knowledge Graph

With the rapid development of social economy and the improvement of people’s material living standard,a growing number of people travel by air.And with that comes is uncivilized behavior of passengers,which has seriously disturbed the order and safety of civil aviation.Accurate matching of uncivilized passenger information and behavior monitoring analysis are one of the important goals of civil aviation research.Through entity identification and relationship extraction of information in passenger records,knowledge map can be constructed to provide labels for user portraits of uncivilized passengers in civil aviation,and then punishment and service restriction can be implemented.Based on the characteristics of civil aviation and the task of identification and classification of uncivilized passengers,this paper studies the construction method of knowledge graph of uncivilized passengers in civil aviation.The following work is mainly done:1)An entity recognition method for unstructured data(Tag + Bi-LSTM + CRF)is proposed.Considering the civil aviation passenger text record contains more than one entity,the entity model appears in the sentence semantics may contain useful information,to embed the character named entity recognition task BIOES mark embedded in series and embedded,position and combination of both short-term and long-term memory network and conditional random field,this paper proposes a new method for entity recognition.2)A neural network model for entity relationship extraction(R-Bert-CNN)is proposed to provide a theoretical basis for the establishment of knowledge graph research.In this model,the entity level information is integrated into the pre-training language model,and the convolutional neural network is used to extract the semantic information at the sentence level,which can better capture the semantics of the sentence and the two target entities.The experimental results show that 89.51% of F1 value is obtained without relying on any natural language processing tools,which effectively improves the efficiency of relation extraction.3)To construct the knowledge map of civil aviation uncivilized passengers.Based on the analysis of civil aviation data,the definition of entity and relation is given,so as to construct the domain concept layer.Knowledge is extracted from unstructured data,and knowledge fusion is realized by using entity dis-ambiguities method.Knowledge is stored by using Neo4 j graph database,and knowledge map of civil aviation uncivilized passengers is constructed.