This document summarizes a lecture on graph algorithms and PageRank using MapReduce. It discusses representing graphs in MapReduce, performing breadth-first search, finding shortest paths, and calculating PageRank through an iterative process of redistributing PageRank values along edges in the graph. The PageRank algorithm is broken into phases that map nodes to PageRank fragments, reduce to calculate new PageRank values, and iterate until convergence is reached. While MapReduce has limitations for iterative algorithms, this approach allows processing graph partitions in parallel through multiple MapReduce jobs.