Join us for the Honeynet Workshop 2024: May 27th–29th, Copenhagen, Denmark

Introducing Acapulco: Building Clustered Parallel Coordinates Graphs from HPFeeds data

20 Aug 2012 Hugo Gascon clustering dbscan hpfeeds kmeans parallel-coordinates splunk
…and the summer is over. During the last three months I have tried to make sense of the highly unstructured data set that comes from merging the data streams of several hpfeeds channels. I have had to learn the inner workings of Splunk, their SDKs, the D3.js graphic library and explore different machine learning frameworks and clustering algorithms. Today I am presenting the first release of Acapulco, a tool to find and display clusters of meta-events built from different types of hpfeeds events within a parallel graph, one of the best ways to represent multidimensional security data in a single visualization.