The increasing amount of Linked Data on the Web can be reused to facilitate numerous applications. One of the first steps is to explore these structured data to determine whether there is relevant information. Since an entity-centric model closely reflects the real world, it provides an intuitive way to explore Linked Data. However, large numbers of linked entities and high diversity of links between entities, often make it difficult for users to understand the overall structure, as well as find the entities of interest quickly for further exploration. In this paper, we present a link pattern discovery approach to facilitate entity exploration. Link patterns describe explicit and implicit relationships between entities and can be used to categorize linked entities. On top of link patterns, we construct a hierarchy to allow exploration of linked entities in a hierarchical multiscale fashion. To lighten users’ exploration burden further, we select top-k link patterns from hierarchy as navigation options. The proposed approach is implemented in a Linked Data browser called SView. We compare it with two conventional Linked Data browsers by conducting a task-based user study. The experiment results show that our approach provides effective support for entity exploration.