Shape Decision-Making in Map-Like Visualization Design Using the Simulated Annealing Algorithm

Abstract

Semantic information without spatial characteristics can also be visualised in the form of a map. This map type is usually called map-like visualisation; it is not a representation of a real geoentity but borrows the map metaphor. In current map-like visualisation works, the manipulation of the cartographic process is poor, and there is a considerable amount of randomness and uncertainty in the map results. For instance, in the process of converting semantic objects to map objects, the map shape is usually uncertain and has many possibilities. Frequently, however, we need shapes to follow certain principles, such as the principles of being close to a certain shape and having a simple geometry, to facilitate recognition and mapping. To achieve a balance between uncertain shapes and design needs, this study enhances the realisation of the cartographer's design idea by improving the controllability in the cartographic process. An optimisation technology, the simulated annealing algorithm, is introduced to control the determination of shape to improve mapping efficiency. Experiments based on real data show that based on this method, the map-like representation not only creates a whole outline close to the pre-defined shape but also realises internal regions with a simple shape.

The State of the Art in Map‐Like Visualization

“…These approaches thus support reading off the spatial distribution of a variable. Moreover, the terrain of topographic maps is imitated by encoding equal levels of hierarchy in an appropriate color scale that represents low‐value regions as green fields and high‐value regions as mountains in the landscape [GJ13, GJKM13, BAYP*15, SSL*17, XAA18, AXY*19, Big20]. Often, a blue color is used for the background [HGK10, PBA11, MKH12, HKV12, GJ13, BAYP*15, SSL*17], which lets areas appear like islands in an ocean.…”

“…Similar to political maps for example, some approaches assign individual colors to each area in the visualization, for example visually differentiate clusters detected in the data [Ves99, PBA11, MKH12, AHL*13, GHK14, YBA15, VMP16]. Other approaches use the fill color of regions similar to choropleth maps, which encode quantitative attributes with colors from a continuous color scale [VMP16, XAA18, AXY*19]. Alternatively, color can be used to symbolize a value distribution over a continuous area [Ves99, MSS01], which gives the impression of isopleth maps.…”

Graph drawing using Jaya

“…Additionally, Simulated Annealing was employed to tackle the challenge of creating map-like visualizations, intending to foster cognitive mapping between the map vehicle and the target ontology through psychological association [ 15 ]. The authors suggest using a Simulated Annealing algorithm to guide the decision-making process for shape determination, thereby enhancing mapping efficiency.…”

A Multi-Scale Virtual Terrain for Hierarchically Structured Non-Location Data

“…The current map-like visualizations take either polygons or virtual terrains as the information carriers. The former often simulates the administrative map in the form of polygon nesting [3,4]. The latter simulates natural landforms using landscape metaphors [5].…”