Research On The Geographic Name Information Retrieval Method In Public Serving Platform Of Geographic Information Service

Geographic name is a language code assigned to the particular area geo-entities, and language identification used to distinguish other specific geo-entities. Information of geographical names is the most common type of public information, not only closely related to peoples’ daily lives, but also is the indispensable basic information resources in state administration, economic development, and domestic and foreign exchanges. Since information of geographical names contains text attribute information and rich spatial and temporal information, it becomes the important medium to connect spatial information and non-spatial information, and plays an increasingly important role in the construction of the domestic public serving platform of geospatial information services of digital city.Since information of geographical names contains text attribute information and rich spatial and temporal information, it becomes the important medium to connect spatial information and non-spatial information, and plays an increasingly important role in the construction of the domestic public serving platform of geospatial information services of digital city. And many problems are aroused by the widespread use of traditional spatial query in the public serving platform ofgeospatial information services, such as the tow retrieval efficiency, tow matching accuracy and the limited amount of the search results information. To solve these problems, and in the analysis of the classification of geographical names in public serving platform of geospatial information services and the defects of method to search the traditional geographic names, this thesis proposed the method of information of geographical names retrieval in public serving platform of geospatial information services in the use of domain feature and Lucene. And this thesis also took the geospatial data in exchange platform of Zhejiang geographic information as the experimental subjects to verify the retrieval quality and efficiency of the method. The research contents and results in this thesis are as follows:(1) Research on the management method of domain feature and index of geographical names:aiming at meaning and classification of the multisource geographic name in platform, this thesis built the domain feature library and its corresponding management methods based on the statistical methods of high-frequency words in information entropy and studied the method to build and maintain the full-text index for multi-level geographic names based on Lucene in platform.(2) Research on the retrieval and sorting methods of geographical name based on Lucene and domain feature:This thesis studies the technology framework of retrieval method of geographical name based on Lucene and domain feature, and on the use of segmentation analysis on the base of domain feature library, discussed the query expression parsing methods oriented to the domain feature library; This thesis optimized the sort results of information of geographic names according to the comprehensive evaluation index composed of text correlation and location correlation.(3) Prototyping and verifying the retrieval system of information of geographic names:This thesis took the geospatial data in exchange platform of Zhejiang geographic information as the experimental subjects and designed the information of geographic names retrieval prototype system based on the domain feature and Lucene. On the analysis of the two common methods, SQLobscure inquiry method and oracle full-text index method, to retrieve information of geographic names in the platform, this thesis validate the effectiveness of the proposed approach.