The Framework Model And The Data Index Of Heterogeneous Cooperative GIS System

Cooperative GIS(CoGIS)is to combine CSCW with GIS.The CoGIS supports a group of collaborators to edit,design,or make decision by the shared GIS information at the same time from geographically dispersed sites connected via the Internet.The essence and goal of the CoGIS technologies is to use the computer and the communication technology to establish a collaborative work environment based on the geography information.Besides some common requirements with the CSCW(Computer Supported Cooperative Work)systems,the GIS has some features such as multiple editing data resource and long-during transaction which are not considered in the CSCW area.To establish a coordinated GIS platform,we must study and design a middleware framework model which is proposed based on component to adapt the GIS systems.Firstly,this thesis analyzes the current development,the architecture and the status quo of the CoGIS.The key techniques in the CoGIS such as multiple editing data resource,long-during transaction and versions handling mechanism are also researched carefully.Secondly,this thesis discusses the system architecture model of the CoGIS,and also presents a replicated system model which supports the shared awareness.One of the key techniques in the CoGIS is real-time shared awareness that supports person-to-person effective cooperation and encourages their creativity or brainstorm to make decisions.Furthermore,the above model is introduced into heterogeneous single-user GIS applications,and a middleware framework model is proposed based on component with a discussion of related interface design of all parts of the middleware,which is the base for building the CoGIS systems.Thirdly,for the spatial data index in the coordination of heterogeneous middleware,this thesis adopts R-Tree to design the bottom level of the local data index.As defined in the paper presented by Guttman,R-Tree is a hierarchically structured high-balanced tree consisted of non-leaf node and leaf node.The minimum out-connected rectangle of real data object is stored in leaf node.The non-leaf node is formed according to the accumulation of its low-level nodes which contains all the out-connected rectangle.This thesis gives a specific and detailed exposition of the R-Tree design,the data structures and the algorithms.Finally,this thesis uses R-tree indexing to establish an index to a GIS system and has implemented the local spatial data inquiry.At the end this thesis makes the analysis and the forecast on the global space data index.