Research On Key Techniques Of Keywords Extraction From Chinese Civil Aviation Safety Information

The development of civil aviation industry results in explosive growth of the safety information,which causes a problem of digging and analysis.In order to solve this problem,it is immediate to design and develop a system for digging safety information effectively.This paper proposes a model of digging keywords based on Naive Bayesian model,after studied the technology of digging keyword and the feature.Accuracy rate and recognition rate of civil aviation is significantly improved,compared with traditional algorithm in performance experiment.The effect of digging the information with five indexes of extracting information is better than the method with three indexes.Then the application of extracting keyword in classification of the safety information in civil aviation and similarity calculation for theme is studied.By taking the keyword extracted in the experiment as feature terms,the space dimension is reduced and the process of feature calculation is simplified,in the same time,the performance of this extracting method is remained.The experiment result shows that the method,which with improved weights algorithm,increases the performance of extracting keyword drastically and the effect of classification for different kinds of safety information in civil aviation,compared with traditional algorithm.The improved algorithm of theme similarity in safety information of civil aviation based on VSM model is advanced.This method avoid some defects in traditional method,such as too many terms of feature,complex of calculation,redundancy feature information,and it can calculate the theme similarity of safety information in civil aviation quickly.So the algorithm advanced in this paper produces a new idea and method for management the safety information in civil aviation.