Research On Key Theories And Methods Of GIS Knowledge Engineering

In the last50years, there have being remarkable achievements in many fields ofGeographic Information System (GIS), such as research on fundamental theory andtechnology, development of common platform and thematic systems, supply ofgeographical information services, and so on. GIS has been a fundamental implementwhich be directly related to socio-economic development and people’s daily life. Itindicates that the development of GIS has entered the stage of public-orientedservicing. The growing demands of GIS knowledge is more and more obvious indevelopment trends of GIS include that the applications of GIS are increasinglypopular and the functions of GIS are increasingly intelligent. Not only all walks of lifeneed to share theoretical and methodological knowledge have been had in GIS，butalso intellectualization of GIS functions such as spatial analysis, geographicalsimulation and optimization need GIS knowledge. Under the background, GISknowledge engineering is introduced and becomes a new research direction in GIS.The important research task in GIS knowledge engineering at present is to learntheories and methods of knowledge engineering (KE). There are great theoreticalsignificance and application value in researches on key theories and methods of GISknowledge engineering in the paper.In the paper, researches on key theories and methods of GIS knowledgeengineering are closely carried out by two main ways: researching on theory andtechnology and researching on engineering and systems. In researches on theory andtechnology, there are three subjects as follows: Introducing and systematicallyillustrating three theories supporting researches on GIS knowledge engineering;Defining clearly concepts and structures of four types GIS knowledge and buildingthe hierarchy of GIS knowledge; Researching systematically technologies ofacquisition, representation, storage, and management of GIS knowledge. In researcheson engineering and systems, there are three practices as follows: Summarizingsystematically the situation of researches on GIS knowledge engineering and definingthe aim of researching GIS knowledge engineering at the present stage; Realizing key technologies of GIS knowledge engineering and constructing a prototype of GISknowledge base; Developing natural language spatial query system (NLSQS) which isan implication system based on GIS knowledge and verifying the scientificness andavailability of key theories and methods of GIS knowledge engineering usingNLSQS.There are there innovative works of researching on key theories andmethods of GIS knowledge engineering in the paper as follows: Raising three baseproblems in researching process of GIS knowledge engineering which are “howpeople cognize geographic space”, how knowledge is generated and how intelligentbehavior is produced by knowledge” and “whether computer can representknowledge and cognitive behaviors of people, and if computer can do, how it do”. Inorder to settle those problems, three important theories are introduced which arespatial recognition, epistemology and symbolism. The scientificness of researching onGIS knowledge engineering can be ensured by researching those problems andtheories. Building the framework of GIS knowledge hierarchy from a combinativeperspective which combined by functions of GIS and types of knowledge. Theframework is not only laying the conceptual foundations for researching theories andmethods of GIS knowledge engineering, but also providing ablueprint for long-term dynamics research on GIS knowledge engineering.In allusion to special characteristics of GIS knowledge, researching knowledgeengineering methods on various types of GIS knowledge. The research way not onlyensure the clarity of research on GIS knowledge engineering, but also enhance thepertinence and availability of methods of GIS knowledge engineering.There are several outcomes of key theories and methods of GIS knowledgeengineering in the paper using which not only GIS knowledge can be shared, but alsointellectualization of some GIS functions can be realized.