Abstract
dc:description.abstractStretched by increasing demand and decreasing budgets, like many local authorities, Leeds City Council have turned to geodemographics to support data-led decision making. As per the current trend for transparent research and policy development, the literature increasingly recommends open geodemographics for use in the public sector. However, the only open classification currently available, the 2011 OAC, which is derived at national level from decennial census data collected in 2011, has proven ineffective at identifying some of the unique multivariate local phenomena. This thesis generates a new framework for a public sector focused place-specific geodemographic classification for Leeds. Primarily, the study introduces and explores the impact of making a methodological shift in geographic extent from national to local level. Secondarily, the research extends beyond traditional decennial census input data to include novel data from open and public sector sources. To support this extension, the work also investigates the potential of several Feature Extraction and Feature Selection methods to intelligently reduce the set of candidate input attributes by identifying those most capable of generating meaningful classification outputs to suit public sector requirements. This thesis demonstrates that there is both scope for generating locally specific classifications with novel administrative data, and benefits to be gained, particularly in terms of identifying locally specific phenomena capable of enriching public policy and decision making processes. It also makes a strong argument for an increased emphasis on incorporating intelligent variable selection processes into geodemographic classification development. The work has been completed during an ESRC collaborative PhD studentship in partnership with LCC and TransUnion. All developments made have primarily considered the needs of a public sector end-user, however, the outputs are transferrable and applicable beyond the public sector. Moreover, transparency and reproducibility has been prioritised to enable and support replications in other cities with similarly available data.
Degree
thesis:*- Name dc:type.qualificationname
- Ph.D
- Level dc:type.qualificationlevel
- doctoral
- Grantor dc:publisher.institution
- University of Leeds
- Year dc:date.issued
- 2021
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Otley, Amanda Leigh
- Advisors dc:contributor.advisor
-
- Morris, Michelle
- Newing, Andy
- Birkin, Mark
Rights
- Language dc:language.dcterms_iso639_2
- eng
Identifiers
dc:identifier.*- Identifier
- uk.bl.ethos.829689
- Dc Identifier Grantnumber
- ES/J500215/1