Place-based GIS allow digital systems to provide a human-centred representation of the geographic world, by complementing traditional spatial representations with the notion of human meaning. An emerging question of such an integration and hence digitization is the level of formalization and generalization that the human meaning can undergo, along with the pragmatic value of associating informal and vague constructs with the formal and precise environment of a Geographic Information System. We propose a function-based model of place, which depicts place as a space ascribed with functionality. The model treats place as a topological graph of spatial entities that enables a set of functions, which in return define functional spaces. Furthermore, utilizing the idea of functional space we suggest a practical application of a Place-based GIS, such as function-based search of space, that is demonstrated using the example of a shopping area. Future research includes the extension of the model by associating place with purposes and emotions, automated generation of composition patterns of place and extraction of function-based data.
A geographic information system (abbr. GIS) represents
geographic information utilizing various forms of data
structures. From “a puzzle of polygons to a sandwich of data
layers” [
This integration forces researchers to focus on the essential difference between space and place. It stems from the intuitive mental ability of creating symbols by ascribing meaning to the physical constructs. With respect to our discussion about space, humans create symbols, which are the places, to assign context to space. The main research question that emerges focuses on the limitations of formalizing the context of space. Particularly, which parts of the human context ascribed to space facilitate generalization and allow formalization; furthermore, what is the pragmatic value of such an integration within GIS, in terms of practical applications and solutions. This work proposes a multidimensional definition of place as functional space and suggests a pragmatic application that utilize the gradual transition from human context to pure spatial representations displayable in GIS platforms.
The rest of this document is organized as follows. The next
section introduces the leading definitions of place and
continues with a brief literature review of the existing
methods that associate space with place. Afterwards, a
definition of place is given that is derived from the theory of
the Object of Discourse [
The academic definition of the concept of place is a research
problem that has puzzled scientists for many years. Relph [
Since human meaning is broad, purely subjective and disruptively vague, any attempt of formalizing it, should be compliant to reasonable abstractions. These facilitate the distinction of indicators that outline the human meaning from those that are not worthy of modelling and determine whether inter-subjectivity can be applied. Adhering to such abstractions facilitates the declarative formalization of meaning. This simplified version of meaning is referred to as context of space and affords semantics representation.
There are two notable directions when conceptualizing and
formalizing place: either infusing spatial representations with
semantics or projecting semantics on space. A leading
approach of augmenting space with semantics is the
objectification of space [
In the opposite way, digital gazetteers [
Most of the aforementioned methods do not fully utilize the expressive power of place. The first three methods associate space with simple semantics, in the sense of properties, which does not always reflect the given human context. On the other hand, the affordance-based model sufficiently approximates the context of place. However, affordances are perceptioncentric and provide limited and individual-driven knowledge. This limits the model's capabilities on defining whether space affords a final goal, which, in turn, impedes the model’s operationalization. 3
The objective of this work is to propose a model that facilitates an adequate conceptualization of place, which allows its representation using a rigid, digital alphabet. This consequently allows the integration of place into GIS platforms. Considering the complexity of the problem, the initial research question is analysed into several coarsegrained challenges. Before listing and addressing the individual challenges, a brief introduction of the theory about the object of discourse is provided, which is the basis of the proposed approach.
An object of discourse is “whatever people can talk about
regardless of its nature” [
The most crucial challenge when it comes to place modelling is the conceptualization of the spatial context. Inspired by the argument that place is an object of discourse, we address this by assuming place is space that offers particular functionality. This allows a more sophisticated and formalizable view of the spatial context that goes beyond simple properties, such as names or attributes. In addition, functions allow operationalization by facilitating objectivity. They provide an inter-subjective understanding on how places operate as a system, rather than following affordances-driven questions such as how the place can be interacted with, which entails individual spatial perception. It is worth noting that functionality expresses only a subset of spatial contexts. Places can be more complicated when they are related to emotions, experiences and so on. For that reason, we list the following assumptions: the proposed model represents only places that exist in the real world, are marked by human intervention and are designed for certain goals.
The next challenge that arises is the encapsulation of the context of place (that is, functionality) in a system of entities that affords realization and spatial representation. This is addressed by following the principles of the object of discourse based on which, functions are enabled by a particular spatial organization. Our approach follows this idea by introducing the composition of place. More specifically, composition suggests a network-based view of place. Every vertex is considered as an entity, denoted as component, and every edge resembles possible associations between components. The components depict physical entities with ascribed properties and rules that offer a generalized description of their potential geometry. The components are associated with spatial relations revealing a possible topology. This topology implies a spatial organization that enables the functions of the modelled place to be offered.
The final challenge that needs to be addressed is the projection of place on space. This can be addressed by utilizing the dual nature of the composition of place that was described above. With respect to the object of discourse, the level of properties realizes a composition by assigning values and creating a tangible representation of the individual object on space. Particularly, the composition of a place is regarded as a blueprint. This includes descriptions of required and optional components along with their topological rules that enable the functions, which form the context of the place under consideration. Since the components are equipped with their geometrical descriptions, they can be populated with real data and spatial objects. As long as these components are spatially organized based on the composition rules, it is then possible to assign the initial context on space and hence project the place itself on space.
Considering all the above, we propose a multi-faceted definition of place incorporating the dimensions of spatial properties, composition and functions. The dimension of spatial properties describes place as a semantically enriched spatial object with ascribed properties and geometry-related information. The dimension of composition describes place as a system in the sense of a topology network. This graphoriented representation resembles the spatial organization of the components that constitute a place. Finally, the dimension of functions provides a sense of context by depicting the set of operations that a place can offer.
For the proposed model to support domain independence, its
formalization should be flexible, reusable and extensible. This
can be achieved via an ontology design pattern [
The composition pattern of the shopping centre (Figure 2a) is used for the functional infusion of an area in Santa Barbara County. Particularly, this example demonstrates the search of place and specifically the function-based search of space. The objective is to locate all places that offer the functions of a shopping centre using OpenStreetMap data. Figure 3 shows the results of the procedure along with the corresponding query of “shopping centre” using the Google Maps platform.
There is a satisfying similarity between the two result sets. However, as opposed to the traditional place name search, the functional infusion includes all places that adhere to the same rules, even if they are not registered as shopping centres. As a result, there are places that are not included in the results of the Google Maps query, such as the area close to the University Campus and the place “The Shop”. In addition, functional infusion is not limited to locating a place but also provides an estimation of its spatial extent.
The current state of this work focuses on the detailed
formalization and evaluation of the dimension of composition.
Particularly, we emphasize on specializing the components’
geometric descriptors by including features such as scale,
fuzzy boundaries and image schemas [
An interesting direction of future work is the extension of the model of place to the planes of intentionality and emotions. More practical future directions include the automation of the following tasks: (a) acquisition of functionbased data; (b) extraction of functions based on purposes that people assign to places; (c) extraction of place composition patterns based on a set of functions.
The presented work is framed within the Doctoral College GIScience (DK W 1237N23), funded by the Austrian Science Fund (FWF)