A method for construction of knowledge graph for land cover processing services
With the advancement of computer and network technologies,a with the advancement of computer and network technologies,a processing services has been published and However,due to the lack of effective correlation among these surface cover processing services,users often find it challenging to select a service that best meets their specific needs,thereby hindering the fulfillment of personalized service selection requirements.Knowledge graphs offer significant advantages in managing heterogeneous data from multiple sources by representing knowledge nodes and their semantic relationships in a structured manner,transitioning from data interconnection to knowledge interconnection.Employing knowledge graphs to semantically represent services,correlate service knowledge,and construct a networked service knowledge graph for surface coverage processing presents a viable approach to realizing personalized service demands.This paper employs semantic web technology and ontologies to standardize terms and concepts within service descriptions,utilizing a unified semantic model to describe services and thus address the issue of multi-source heterogeneity.Initially,domain knowledge pertaining to land cover treatment services is organized based on relevant literature,establishing a unified conceptual framework for describing treatment service knowledge.Subsequently,guided by this conceptual framework,a corresponding template wrapper is constructed to extract service attribute information,compute cosine similarities between service descriptions,analyze semantic relationships among service entities,and develop the service knowledge map for ground coverage processing.Finally,a prototype knowledge map system for land cover processing services is developed,enabling users to search for required services using keywords.Following the conceptual framework for land cover processing services,a Python-based template wrapper is created to extract attribute information from 1109 services.By calculating cosine similarities between service descriptions and analyzing semantic relationships,a knowledge graph comprising 11863 attribute nodes and 11012 relational edges is formed.Using frameworks such as Django and Vue,a knowledge map application system for surface coverage processing services is established.The system selects services with semantic similarity scores above 0.7 as matching results for user choice by comparing user search keywords with service descriptions.Additionally,the interrelationships among services provide users with a broader selection,catering to their needs for personalized service selection.Given the challenge of effectively correlating multi-source heterogeneous land cover processing services,which impedes personalized selection,this study explores the construction and application of knowledge graphs in the realm of land cover processing services.Concrete application examples demonstrate that the knowledge graph for land cover processing services can effectively satisfy users'needs for personalized service selection.Future research should focus on expanding data sources,enhancing the comprehensiveness of the land cover processing service knowledge graph,and further improving the functionalities of the prototype system,aiming to intelligently construct land cover processing service chains according to user requirements.
NETL
NSTL
万方数据
