Research On Key Technology Of Geographic Entity Information Acquisition Based On Internet

With the explosive growth of information resources on the Internet,all walks of life data floodes in the sea of the Internet.Geographic entity information collection mainly relies on the national geographic information census,field measurement and investigation,which is timeconsuming and laborious.It is difficult to meet the demand of geographic entity information rapid gathering and updating,and the rich and diverse information on the Internet provides a strong support for the rapid acquisition of geographic entity information.Based on the research about the theory and technology of Internet access to geo-physical location data and geographic entity attribute data,this paper establishes a prototype system of Internet-based geographic entity information collection.The specific contents of the paper are as follows:1.On the basis of the analysis of geographic entities collection's background,the importance of collecting the geographic entity information based on the Internet is pointed.The Internet distribution status of Web geographic information and the research status of the Web geographic information collection are introduced.Based on the information of geographic entities in the text information,the research status of information extraction is analyzed.This paper summarizes the related concepts of geographic entities,and compares the similarities and differences between geophysical information and basic geographic information,POI information.The related concepts,methods and methods of information extraction are introduced.This paper introduces the spatial data and attribute data acquisition of geographic entities,and puts forward the technical scheme of obtaining the information entity of the geographic entity,based on the concept of geographic entity and information extraction.2.In view of the access of the geographic entity spatial data from the Internet,this paper proposes a global geospatial entity data acquisition scheme based on POI.We use Baidu map and AutoMap API to implement a quick acquisition of all POIs in arbitray region in China.We use OSM to conduct research and experimentation in foreign regions and propose two acquisition schemes for extracting POI data considering data format conversion and <key,value> in OSM data.Finally,by combining the different characteristics of the three data,merging the Baidu map and the AutoMap POI data,and transforming the OSM data the location data of the geographic entity based on the POI data is generated.3.This paper studies the geographic entity attributes extraction of semi-structured text and unstructured text,which is based on the structured data acquisition of geographic entity attribute and the characteristics of geographic entity attribute in the encyclopedic information.The former is geographic entity attributes extraction based on an information box,which put forward through the theory of extracting geographic entity attributes through attribute normalization and text rules matching.The latter uses the machine learning theory and proposes a weak supervisory method based on conditional random field extraction.Labelling the unstructured text automatically using geometric entity attribute getting from the former method and generating the training set,then this paper use conditional random field model generate geographic entity attribute feature model,and extracts the attribute of geographic entity in unstructured text predictably with this model.Finally,the related experiments were designed and verified aimed at the above two methods.4.According to the realistic demand of geographical entity information acquisition,this paper summarizes the technical and achievements of the above-mentioned geographic entity location data acquisition and attribute data acquisition,which base on the rich and diverse basic Web geographic information service,and designs and implements a prototype system of geographic entity information collection on basis of the Internet,which applies the above research results.At the same time,the study achieves the integration of the geographic entity information collection based on the Internet.