Exploring multi-granular documentation strategies for the representation, discovery and use of geographic information

This thesis explores how digital representations of geography and Geographic Information (GI) may be described, and how these descriptions facilitate the use of the resources they depict. More specifically, it critically examines existing geospatial documentation practices and aims to identify opportunities for refinement therein, whether when used to signpost those data assets documented, for managing and maintaining information assets, or to assist in resource interpretation and discrimination. Documentation of GI can therefore facilitate its utilisation; it can be reasonably expected that by refining documentation practices, GI hold the potential for being better exploited. The underpinning theme connecting the individual papers of the thesis is one of multi-granular documentation. GI may be recorded at varying degrees of granularity, and yet traditional documentation efforts have predominantly focussed on a solitary level (that of the geospatial data layer). Developing documentation practices to account for other granularities permits the description of GI at different levels of detail and can further assist in realising its potential through better discovery, interpretation and use. One of the aims of the current work is to establish the merit of such multi-granular practices. Over the course of four research papers and a short research article, proprietary as well as open source software approaches are accordingly presented and provide proof-of-concept and conceptual solutions that aim to enhance GI utilisation through improved documentation practices. Presented in the context of an existing body of research, the proposed approaches focus on the technological infrastructure supporting data discovery, the automation of documentation processes and the implications of describing geospatial information resources of varying granularity. Each paper successively contributes to the notion that geospatial resources are potentially better exploited when documentation practices account for the multi-granular aspects of GI, and the varying ways in which such documentation may be used. In establishing the merit of multi-granular documentation, it is nevertheless recognised in the current work that instituting a comprehensive documentation strategy at several granularities may be unrealistic for some geospatial applications. Pragmatically, the level of effort required would be excessive, making universal adoption impractical. Considering however the ever-expanding volumes of geospatial data gathered and the demand for ways of managing and maintaining the usefulness of potentially unwieldy repositories, improved documentation practices are required. A system of hierarchical documentation, of self-documenting information, would provide for information discovery and retrieval from such expanding resource pools at multiple granularities, improve the accessibility of GI and ultimately, its utilisation.EThOS - Electronic Theses Online ServiceGBUnited Kingdo