| In order to fully mine civil aviation safety data to provide useful information for civil aviation safety decision-making.It is necessary to determine the competency index system of civil aviation safety information management personnel that meets the requirements of big data environment,and construct an accurate and efficient competency-post matching evaluation model of civil aviation safety information management personnel,so as to provide a scientific basis for the recruitment,selection and promotion of civil aviation safety information management personnel.Firstly,the content of competency of civil aviation safety information management personnel is defined.On this basis,using the common data life cycle and basic information processing(DIKW)hierarchy structure for reference,the concept model of civil aviation safety information management personnel competence is constructed.Based on the conceptual model and literature analysis,the competency index system of civil aviation safety information management personnel is determined,including data science knowledge,civil aviation safety management knowledge,the competency to identify hazard and collect civil aviation safety data and other 7 first-level indicators and 25 second-level indicators.Then,based on Bayesian optimization BP neural network,this paper constructs a competency-post matching evaluation model of civil aviation safety information management personnel,and carries on the training and verification of the model.The results show that,compared with random search optimization algorithm,the root mean square error of Bayesian optimization algorithm is reduced by 0.02.Compared with the single BP neural network,the root-mean-square error of the model constructed in this paper is reduced by 0.053,and the algorithm time is reduced by 2.25 s,which obviously improves the accuracy and speed the single BP neural network in the evaluation of competency-post matching.Finally,5 civil aviation safety information management personnel are taken as examples to carry out an example verification.The results show that the competency-post matching degree of No.1 evaluator is "general match",No.2 evaluator is "very match",and No.3-5evaluator is "relatively match",which is consistent with the actual situation. |