This document discusses spreading activation as a method for searching associative networks and semantic networks. Spreading activation is based on quickly spreading an associative relevancy measure over a network. The method is presented as a general framework and class of algorithms that can be applied to problems involving large multidimensional networks. Key aspects of spreading activation algorithms include initializing the method, running iterations where activation spreads from node to node, and outputting the results. Spreading activation can be used for tasks like soft clustering and fuzzy inferencing on networks.