In-depth Analysis And Research On Information Report Of Airline Employees Based On Knowledge Map

Civil aviation safety information is an important basis for discovering potential safety hazards,controlling risks and improving the safety of civil aviation.Mandatory reporting is a well-deserved source of information to improve the safe operation of aeronautical systems and airline employee reporting is an important part of mandatory reporting information.In order to dig out the potential safety hazard in the employee report information and establish the knowledge map network between the causes and consequences of many kinds of incidents to provide the basis for data analysis for the safety risk analysis.In this paper,the 1227 unsafe incidents reported information of four types of landing events were taken,such as hard landing,aborting approach and take off,interruption of communication,and less than interval,caused by the flight crew collected in 2015 and 2016 by CAAC Safety Information Network.Using the method of knowledge map network analysis,investigate the time,the stage of flight,the causes and consequences of occurrence of the incident in the incident report due to the cause of the flight crew.Analyze the part of speech in the extraction result and combine artificial error vocabulary to obtain the result and use keyword extraction technology to extract keywords and phrases from event description as reference for selecting node of map network.According to the classification of incident reports,keyword extraction,the construction of knowledge map network framework,the process of determining the relationship between nodes and node layers,the construction of sub-knowledge map network for landing events and other three types of event sub-knowledge map networks.Then,a crew-centered knowledge map network of airline employees' report information is established.Finally,a prototype system of employee report information based on this method was constructed by using knowledge map analysis method,which provides an idea for the intelligent analysis of report information.An empirical analysis of the knowledge map network for unit causes based on a hard landing event.The results show,according to the knowledge map network to analyze the four types causes by crew,it can quickly get the information about the causes and consequences of the "human-machine-environment" involved in the incident so that the event evolution becomes clear.And according to the knowledge map network's analysis of events of a certain type reason used the DECIDE model analysis method to find out the influence of the operation of crew when the event occurs and to make the corresponding control measures.In addition,according to the reason nodes,consequence nodes and basic information nodes can accurately trace the events and facilitate the event search.Moreover,the new event causes and consequence node information can be added to the knowledge map network to enrich the content continuously.Due to the limited amount of information in the sample event description used in this paper,there are still some shortcomings in keyword extraction,such as the dependence on the lexical inclusion of specialized dictionaries in terms of the accurate extraction of specialized vocabularies.The greater the amount,the accuracy and amount of extraction will increase accordingly,so the selection of nodes using keyword extraction results will also bring deviation,this area can continue in-depth research and analysis.