关键词: 民航突发事件, 文本信息, 领域本体, 概念获取, LDA模型

Abstract: Aiming at the automatic updating of the ontology of emergency management of civil aviation emergencies, a method of domain ontology based on LDA（Latent Dirichlet Allocation） is proposed. Based on the text information as the data source, the NLPIR adaptive word segmentation and filtering method are used to obtain the candidate term set. The LDA theme model of the domain ontology is designed, and the LDA model training is carried out by Gibbs sampling and the topic is deduced, which realizes the core concept of the domain ontology. The semantic relation recognition rule construction method is studied based on the probability distribution of LDA subject, and the recognition and realization process of semantic relation of concept and its related term are given. The experimental results show that this method can effectively solve the problem of automatic updating of large-scale domain ontology concept, and provides good data support for the sharing and reasoning of cross - media information of civil aviation emergencies in large data environment.

Key words: civil aviation emergencies, text information, domain ontology, concept acquisition, Latent Dirichlet Allocation（LDA）