Datasets in KONECT represent networks, i.e., a set of nodes connected by links. Networks can be classified by their format (directed/undirected/bipartite), by their edge weight types and multiplicities, by the presence of metadata such as timestamps and node labels, and by the types of objects represented by nodes and links.

The format of a network is always one of the following.

• In undirected networks, edges are undirected. That is, there is no difference between the edge from u to v and the edge from v to u; both are the edge {u, v}. An example of an undirected network is the social network of Facebook (Ow), in which there is no difference between the statements “A is a friend of B” and “B is a friend of A.”
• In a directed network, the links are directed. That is, there is a difference between the edge (u, v) and the edge (u, v). Directed networks are sometimes also called digraphs (for directed graphs), and their edges arcs. An example of a directed social network is the follower network of Twitter (TF), in which the fact that user A follows user B does not imply that user B follows user A.
• Bipartite networks include two types of nodes, and all edges con- nect one node type with the other. An example of a bipartite network is a rating graph, consisting of the node types user and movie, and each rating connects a user and a movie (M3). Bipartite networks are always undirected in KONECT.