No abstract

The elements that constitute the literary space can mainly be described by categories such as settings, zones where actions take place and routes along which characters move. Apart from the presentation of the specific locations and spatial distribution of literary places, we are also interested in the spatial pattern such places form, the boundaries that separate the literarily populated from the void regions and the varying density of the literary space. From a GIS point of view, the elements of the literary space correspond to point, line and area data types. However, the established method used to calculate the spatially varying density -the method of density estimation -is restricted to point data only. In this paper, we will present an improved method that is able to estimate the density regardless of the underlying data type. Our approach aims at adopting the typically radially symmetric kernel function to approximate also linear and areal features. We claim that this method treats point, line and area data in a consistent way by taking equivalent density contributions into account. The different steps of the improved method can visually be examined by the accompanying map examples.

No abstract

No abstract