The "Geospatial Analysis: GIS & Agent-Based Models" session at the RGS-IBG Conference included 5 interesting and ongoing research works which are summarized below.
Note:It is also expected to run the Geospatial Analysis: GIS & Agent-Based Models session next year - RGS-IBG 2009.
ABSTRACTSModelling Perceptions of Street Safety to Increase Access to Public Transport; Claire Ellul, Ben Calnan (Cities Institute, London Metropolitan University) In the context of public transportation, “the provision of a permeable public space contributes to an inclusive journey environment” (Azmin-Fouladi 2007). However, when planning or modelling an urban environment, architectural vision and planning principles often take precedence over the way buildings and urban features make people feel. In particular, the identification of specific urban features that contribute towards a feeling of safety and security is not generally considered.
Our research aims to redress this imbalance by providing planners and local authorities with the means to identify potential barriers to the permeability of public space. It is argued that the removal of negatively-impacting features and the resulting increase in perception of safety will increase the use of public transportation.
We present two key outputs of this process. Firstly, we have developed an Index of Permeabilility (IoP) for the urban environment, where each relevant urban feature visible from a specific location has been assigned a weighting (through a process of consultation). This weighting contributes towards the overall index of permeability for the point. Secondly, we present a GIS-based implementation of this index using Isovists (which identify the urban features visible from a specific point), extending the index to create a surface of permeability.
System wide cultural districts: mapping and clustering the tangible and intangible cultural assets for the policy design of the regional clusters in the Veneto Region, Italy; Pier Luigi Sacco, Guido Ferilli (IUAV University), Massimo Buscema, Terzi Stefano (Semeion Research Center)
In previous research carried out by Sacco et al. the notion of system-wide cultural districts has been introduced and analyzed. In particular, system-wide cultural districts are horizontally integrated local clusters of economic activities in which culture plays a key strategic role as a social activator of innovative processes and practices, as well as an attractor of talent and resources, a factor of social cohesion and of networking, and of course as a sector with its own value added.
In other, related research from the same group, Artificial Neural Networks (ANN) techniques have been adopted to investigate to what degree they were able to single out emergent industrial districts of various kinds in selected areas of the Italian territory.
In this paper, we combine this two strands of research in a project carried out under the initiative of the Veneto Region, one of Italy’s outstanding productive regions. In the first phase of the project, the spatial distribution and clustering of all cultural activities and facilities with a non-occasional character has been mapped. This has led us to identify qualitatively a certain number of emergent culture-driven clustering. In the second phase, a battery of innovative ANN techniques has been employed to identify the ‘centroids’ of the cultural clusters and to check to what extent they overlap with the poles of the Region’s overall productive systems.
Finally, an analogous analysis has been conducted for specific cultural sectors – visual arts, performing arts, museums, and so on, to investigate to what extent they tend to gravitate upon specific cultural clusters and to what extent they are useful to define prospective local specializations by means of a specific policy design process.
Revealing the fuzzy geography of an urban locality; Richard Flemmings (Blom Aerofilms Ltd & Birkbeck University of London)The delineation of urban geographical boundaries can be problematic, particularly when unitary authority boundaries do not represent perceived reality. The lack of agreement between perception and the reality of political boundaries, make an urban locality a fuzzy geography. This fuzzy geography can be exploited, for example by estate agents who wish to alter an area to increase property values. By giving such fuzzy boundaries definition, better clarity can be achieved between estate agent and customer.
A method is proposed here that gives definition to the boundary of an imprecise region using the internet as the information source. Kernel density estimation is used to transform geo-tagged internet search results into a continuous surface. This is both compared and combined with a kernel density estimation of relevant Ordnance Survey MasterMap® cartographic text labels. A composite Index of Urban Locality is given to represent the fuzzy boundary of Clifton, Bristol. The resulting continuous surface is graded based on membership. Thus, the extent that a location is within or is not within the urban locality is depicted. The success of this output has been verified using estate agent’s interpretations of the boundary of Clifton. The Index of Urban Locality has also been applied to the region of Bedminster, Bristol, with some success.
Geospatial Modelling and Collaborative Reasoning of Indeterminate Phenomena: The Object-Field Model with Uncertainty and Semantics; Vlasios Voudouris (London Metropolitan Business School & City University London) The need for a conceptually unifying geospatial data model for the representation of geospatial phenomena has already been acknowledged. Recognising that the importance of the data model employed by and large determines what can be done by way of analysis and the methods by which the analysis can be undertaken, there has been some activity in developing unifying data models for geospatial representation in digital form. Some successes have been reported. Nevertheless, progress has been slow, especially at the conceptual and logical levels of abstraction of geospatial data models.
Concepts and ideas from cognitive and perceptual psychology as well as GIScience and GISystems literature are examined within the context of geospatial data modelling and reasoning. Drawing on and combining these concepts, ideas and successes with an empirical approach, this work presents the fused Object-Field model with uncertainty and semantics at the conceptual and logical levels of abstraction.
The purpose of the Object-Field model is to better support the representation and collaborative reasoning of geospatial phenomena, particularly indeterminate phenomena such as town centres. It is shown that many of the concepts required to better represent geospatial phenomena can be derived from a single foundation that is termed the elementary-geoParticle. This serves as the standard for integrating the dual continuous-field and discrete-object data models by means of aggregation
GIS and Built Form: Using Pattern Recognition for Energy Efficiency Models; Donald Alexander, Simon Lannon, Orly Linovski (Cardiff University)Much of what has been written about residential development in the UK relies on anecdotal evidence (Whitehand and Carr 1999). Little ‘on-the-ground’ research has been conducted due to the significant time required for investigating development through building records and other municipal data. A wide variety of research often requires detailed building information that has previously only been obtainable through walk-by surveys or building records. This paper examines alternative methods for determining building age using pattern recognition algorithms.
This model has wide ranging applications including researching urban development patterns, conducting urban design studies and assessing energy efficiency. This paper specifically focuses on the use of building data for energy efficiency studies. Modelling software has been developed to quantify energy emissions but requires detailed information of the built environment and age of buildings (Jones et al. 2000). It is proposed that pattern recognition algorithms can be used to automate the collection of this data from GIS and aerial photos.
To develop this technique, two study areas in Wales were chosen as case studies. These areas were surveyed manually to establish a baseline for assessing the built form characteristics of each development that could be incorporated into the algorithm. This paper will present the results of the development characteristic study, as well as the efficacy of using these to determine the age of dwellings.
Posts
