Big Knowledge repositories, in the form of large ontologies, typically consist of many thousands of knowledge assertions. They have complex network structures consisting of nodes and links. Without some form of comprehension, humans cannot make correct, innovative and creative use of Big Knowledge. Visualization is an important tool for knowledge comprehension, however, the node-link diagrams become overwhelming for Big Knowledge. In order to support comprehension, we have developed methods for algorithmically summarizing ontology content and visualizing the summaries. These methods facilitate gaining an understanding of the 'big picture' of an ontology, which is essential for maintenance and integration into applications. Such a summary is called an abstraction network. Similar to the theory of limited working memory in humans, we assume that there is a limited human comprehension capacity for node-link ontology diagrams. In this paper, we present a visualization scheme that is based on multi-layer, multi-granularity abstraction networks of ontology content, each of which stays below a maximum number of nodes. We demonstrate this visualization scheme on the National Cancer Institute Thesaurus's Neoplasm subhierarchy.