An Agent-Based Variogram Modeller: Investigating Intelligent, Distributed-Component Geographical Information Systems

PhD Thesis


Al Zakwani, A. 2021. An Agent-Based Variogram Modeller: Investigating Intelligent, Distributed-Component Geographical Information Systems. PhD Thesis University of East London School of Architecture, Computing and Engineering https://doi.org/10.15123/uel.89w63
AuthorsAl Zakwani, A.
TypePhD Thesis
Abstract

Geo-Information Science (GIScience) is the field of study that addresses substantive questions concerning the handling, analysis and visualisation of spatial data. Geo- Information Systems (GIS), including software, data acquisition and organisational arrangements, are the key technologies underpinning GIScience. A GIS is normally tailored to the service it is supposed to perform. However, there is often the need to do a function that might not be supported by the GIS tool being used. The normal solution in these circumstances is to go out and look for another tool that can do the service, and often an expert to use that tool. This is expensive, time consuming and certainly stressful to the geographical data analyses. On the other hand, GIS is often used in conjunction with other technologies to form a geocomputational environment. One of the complex tools in geocomputation is geostatistics. One of its functions is to provide the means to determine the extent of spatial dependencies within geographical data and processes. Spatial datasets are often large and complex. Currently Agent system are being integrated into GIS to offer flexibility and allow better data analysis. The theis will look into the current application of Agents in within the GIS community, determine if they are used to representing data, process or act a service.
The thesis looks into proving the applicability of an agent-oriented paradigm as a service based GIS, having the possibility of providing greater interoperability and reducing resource requirements (human and tools). In particular, analysis was undertaken to determine the need to introduce enhanced features to agents, in order to maximise their effectiveness in GIS. This was achieved by addressing the software agent complexity in design and implementation for the GIS environment and by suggesting possible solutions to encountered problems. The software agent characteristics and features (which include the dynamic binding of plans to software agents in order to tackle the levels of complexity and range of contexts) were examined, as well as discussing current GIScience and the applications of agent technology to GIS, agents as entities, objects and processes. These concepts and their functionalities to GIS are then analysed and discussed. The extent of agent functionality, analysis of the gaps and the use these technologies to express a distributed service providing an agent-based GIS framework is then presented.
Thus, a general agent-based framework for GIS and a novel agent-based architecture for a specific part of GIS, the variogram, to examine the applicability of the agent- oriented paradigm to GIS, was devised. An examination of the current mechanisms for constructing variograms, underlying processes and functions was undertaken, then these processes were embedded into a novel agent architecture for GIS. Once the successful software agent implementation had been achieved, the corresponding tool
was tested and validated - internally for code errors and externally to determine its functional requirements and whether it enhances the GIS process of dealing with data. Thereafter, its compared with other known service based GIS agents and its advantages and disadvantages analysed.

Year2021
PublisherUniversity of East London
Digital Object Identifier (DOI)https://doi.org/10.15123/uel.89w63
File
License
File Access Level
Anyone
Publication dates
Online01 Oct 2021
Publication process dates
Submitted01 Sep 2021
Deposited01 Oct 2021
Permalink -

https://repository.uel.ac.uk/item/89w63

Download files


File
2021_PhD_Zakwani.pdf
License: CC BY-NC-ND 4.0
File access level: Anyone

  • 21
    total views
  • 24
    total downloads
  • 21
    views this month
  • 24
    downloads this month

Export as