This repository contains an implementation of Google's PageRank algoritm, implemented using multithreading with the MulticoreBSP library.

Two different implementations of the algorithm are available. One calculates the Google matrix and stores it in the memory as a dense matrix. By doing this, the program can access all of the values very rapidly and make faster calculations. The downside of this implementation is that it has a memory complexity of O(N²). This means that it uses massive amounts of memory for large link matrices.

A second implementation uses sorted linked lists to represents a sparse matrix. Due to the nature of the adjustments made by Google to the hyperlink matrix, it is possible to calculate the Google matrix values using sparse data during execution very rapidly. Performance compared to the dense matrix implementation is lower, but the difference will only show for very large datasets, which would be nearly impossible to process with the dense matrix implementation due to the memory requirements.