Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.565738
Title: A framework for quality evaluation of VGI linear datasets
Author: Koukoletsos, T.
Awarding Body: University College London (University of London)
Current Institution: University College London (University of London)
Date of Award: 2012
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
Spatial data collection, processing, distribution and understanding have traditionally been handled by professionals. However, as technology advances, non-experts can now collect Geographic Information (GI), create spatial databases and distribute GI through web applications. This Volunteered Geographic Information (VGI), as it is called, seems to be a promising spatial data source. However, the most concerning issue is its unknown and heterogeneous quality, which cannot be handled by traditional quality measurement methods; the quality elements that these methods measure were standardised long before the appearance of VGI and they assume uniform quality behaviour. The lack of a suitable quality evaluation framework with an appropriate level of automation, which would enable the repetition of the quality assessment when VGI is updated, renders the choice of using it difficult or risky for potential users. This thesis proposes a framework for quality evaluation of linear VGI datasets, used to represent networks. The suggested automated methodology is based on a comparison of a VGI dataset with a dataset of known quality. The heterogeneity issue is handled by producing individual results for small areal units, using a tessellation grid. The quality elements measured are data completeness, attribute and positional accuracy, considered as most important for VGI. Compared to previous research, this thesis includes an automated data matching procedure, specifically designed for VGI. It combines geometric and thematic constraints, shifting the scale of importance from geometry to non-spatial attributes, depending on their existence in the VGI dataset. Based on the data matching results, all quality elements are then measured for corresponding objects, providing a more accurate quality assessment. The method is tested on three case studies. Data matching proves to be quite efficient, leading to more accurate quality results. The data completeness approach also tackles VGI over-completeness, which broadens the method usage for data fusion purposes.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.565738  DOI: Not available
Share: