The Extraction Of Geospatial Informationfor Traffic Radio Text

With the rapid development of information technology,a large amount of text information related to geographic data is increasing rapidly.In recent years,the in-depth development of data mining technology,which also makes it possible that the text,video and image based geographic data can service for people's lives and production.However,in the face of increasing textual geographic data,it is time-consuming and backward to adopt traditional manual labeling,and it is far from satisfying to the requirements on the working efficiency.Therefore,fast and automatic methods about text information extraction become particularly important.Traffic radio voice text which contains names,address and location information is road traffic information,and described in the form of natural language.This paper selects the Beijing city traffic broadcast voice text information in 2014 as an example,in order to realize the extraction of geographic spatial information in traffic speech text,the idea of Chinese text information extraction is used to carry out experiments and the extracted information is presented visually.The research content of this paper is divided into four parts.1.Traffic voice text data preprocessing.According to the characteristics of traffic voice text,this paper carrys out the preprocessing of the text data and analyzes the data preprocessing,such as word participle tagging,word filtering and so on.2.Traffic voice text word segmentation.After analyzing the conventional word segmentation method,it is found that the conventional word segmentation method can not realize the word segmentation of traffic speech text well.Aiming at the problem of traffic speech word segmentation,this paper proposes a word segmentation method suitable for traffic voice text.3.Construction and recognition of traffic voice text mode.Traffic voice text,although expressed free and free,but the expression model is relatively fixed.Using this feature,this paper constructs the model and further introduces the pattern recognition model,realizing the unstructured form of traffic voice text information to the effective transformation of structured information and improving the speed of text information extraction.4.Visualization of geographic information in traffic voice text.In this paper,the geospatial information extracted from the traffic voice text is presented through the visualization of the tag cloud,histogram,Arcgis dot density and so on,and the visual representation of the geospatial information in the text is realized.