Monitoring the conformance of planning decisions to urban land use policies using Information Extraction and GeoVisualisation

Abstract

In this thesis two existing computer science techniques are used to solve a specific problem in the field of ‘spatial planning’. The problem to be addressed is monitoring the conformance of planning decisions to urban land use policies. ‘Monitoring conformance’ refers to adherence to development plans and must be distinguished from monitoring performance, which looks at whether or not the plan met its objectives. The two computing techniques applied to the problem are Information Extraction (IE) and GeoVisualisation (GV). IE is an approach to the automated processing of text. This thesis demonstrates that the restricted subset of language used in the short texts present in planning applications makes them ideally suited to IE methods. GV is an approach to the interactive analysis of geographical data. Its use was motivated by two factors. Firstly, it is necessary to avoid the assumption of a simple relationship between policy and implementation – many different policies may apply to a particular decision. These may be weighted differently and are open to interpretation. Hence, statistical conclusions, such as ‘there is 80% conformance to policy’, are never drawn. Instead the visualisation leaves the interpretation of the results open to the user. It is through the details-on-demand functionality of visualisation tools that this link to the user’s own background knowledge is made. Secondly, the prototype user interface developed exemplifies the use of GV to explore geo-temporal patterns in the data. This was motivated by the knowledge that policies change over time. Evaluation work is conducted which shows that policy-makers can see reflections of the conformance of decision making to urban land use policies in the GV tool. The computational techniques used have been brought together and applied to the domain in a novel way, which assists in addressing the problem identified. A number of more theoretical questions are also considered along the way.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 3 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.