Distributed randomized pagerank algorithms based on web aggregation over unreliable channels

The PageRank algorithm employed at Google assigns a measure of importance to each web page for rankings in search results. In our recent papers, we have proposed a distributed randomized approach for this algorithm, where web pages are treated as agents computing their own PageRank by communicating with linked pages. Here, we focus on the effects of unreliability in the communication and, in particular, model the random data losses as an outcome of Markov chains. By generalizing the aggregated PageRank computation previously developed, we provide a distributed scheme along with analyses on its convergence properties.