Networkalization of Network–Unlike Entities: How to Preserve Encoded Information

Abstract

More than for twenty years network science with complex networks as its basic component has brought the idea to analyze a wide spectrum of entities through a focus on relations between the actors and has implemented the concomitant powerful instruments of the analysis. Some entities (objects, processes, and data) with their intrinsic web nature might be interpreted as networks naturally. Network ontology of another family, Network–Unlike Entities, e.g. spatial and temporal ones, is severely ambiguous and encounters with tough problems on the way to convert data into networks. We concentrate on separation the properties of data in line with their scale diversity – in the distance, time, and nature and suggested a 3 step algorithm (scale-based technique) to convert Network–Unlike Entities into complex networks. The technique was applied for networkalization of landscape and land use maps representing Olkhon district, Irkutsk region, Baikal Lake territory, RF. It was found that the technique with its coarse-graining and area-like connecting conserves natural information inherent to the entities and imbeds accordingly scale-free and small world properties into output networks, thus making them really complex in their structure.