In the last few years, Linked Open Data sources have extremely increased in number. Despite their enormous potential, it is really hard to find effective and efficient ways for navigating and exploring them, mainly because of complexity and volume issues. In fact, application developers, students and researchers that are not experts in Semantic Web technologies often lose themselves in the intricacies of the Web of Data. We propose to address this problem by providing users with a map-like visualization that acts as an entry point for the exploration of a dataset. To this end, we adapt a spatialization approach, based on cartographic and information visualisation techniques, to make it suitable for Linked Data sets with a hierarchical ontological structure. Finally, we apply our method on DBpedia, implementing and testing a prototype web application that shows a comprehensive and organic representation of the more than 4 million instances defined by the dataset.