Visualizing Library and Information Science from the practitioner's perspective
Yusef Hassan‐Montero, Víctor Herrero-Solana · 2007
The matrix was visually represented as a network display, where each subject is shown as a node, and their co-occurrence relationships as links between nodes. In order to show a clearer and more comprehensible graph, the less significant links were pruned. The pruning method used was Pathfinder algorithm (Schvaneveldt; 1990), with the parameters q=N-1 and r=∞. Also, Kamada & Kawai’s algorithm (1989) was used to situate the nodes in space. In the final network display (Figure 2), link thickness indicates the relationship weight (co-occurrence frequency), and the node’s label size reflects the number of members.