QuickRank: A Recursive Ranking Algorithm

This paper presents QuickRank, an efficient algorithm for ranking individuals in a society, given a network that encodes their relationships, assuming that network possesses an accompanying hierarchical structure: e.g., the Enron email database together with the corporation’s organizational chart. The QuickRank design is founded on the “peer-review ” principle, defined herein, and an hypothesis due to Bonacich. Together, these premises leads to a recursive ranking algorithm which is scalable, parallelizable, and easily updateable. Moreover, it is also potentially more resistant to link-spamming than other popular ranking algorithms.