Design And Implementation Of Internet Meteorological Disaster Text Multi Labelclassification System

Since the emergence of human civilization,meteorological disasters have always been accompanied by the development of human society.It is a very meaningful subject to study the law of meteorological disasters to reduce the losses and injuries it brings to people.The classification of meteorological disaster text is an important and basic pioneering task.However,there are many kinds of natural disasters,the relationship between disasters is complex,and the number of texts of natural disasters is huge,so it is difficult to classify them quickly and accurately by manpower.Therefore,the application and implementation of text classification technology in the field of disaster has important research significance and implementation value.The purpose of this paper is to apply text classification technology to the field of meteorological disaster,train the classification model for meteorological disaster text,and build an accurate and efficient automatic disaster text classification platform based on this model.This paper uses crawler technology to collect a large number of disaster texts.After text preprocessing,manual screening and category labeling,a disaster text corpus is established.The data in the corpus is used to train the classification model.In the experimental stage,this paper designs a multivariate classification model and a multi label classification model.After a comparative test,two optimal classification models are finally selected and applied in the classification system.After that,this paper builds an interactive system based on FLASK framework,integrates the classification model and database system,and completes the development of Internet meteorological disaster text multi label classification system.Finally,the robustness and stability of the system are tested in detail.This system focuses on the actual landing effect.Some functions of this system have been applied in Provincial Meteorological units,so this paper has high application significance and landing value.