Analysis and clustering of PingER network data

The PingER project was started by the SLAC National Accelerator Laboratory, Stanford, California for the purpose of monitoring end to end network performance. For the last eighteen years PingER has generated an enormous amount of data that has been stored in space separated files. However due to the difficulties faced in retrieving data efficiently, it has been proposed that all the data be put into the form of RDF triples. Interpreting and analyzing such large volumes of data becomes a primary concern. By making using of clustering algorithms new and interesting patterns can be observed in the data sets. Outlier analysis can be performed giving insight to the exceptions occurring in the dataset and analyzing the probable causes of such. Patterns could be observed based on the country to which the data belongs and comparisons can be drawn between the patterns between the different countries.