SALSA and PageRank are algorithms that analyze the link structure of web pages to determine authoritative pages. SALSA models the web as two Markov chains, with one for hubs (pages with many outlinks) and one for authorities (pages with many inlinks). It computes hub and authority scores for pages based on the principal eigenvectors of these chains. PageRank also models the web as a graph, and computes page ranks based on a random walk model where a user randomly clicks on links. It assigns higher rank to pages linked from many highly ranked pages. Both use the concept of a random walk through the link graph to determine important pages.