shr1611/Pattern-Recognition-on-Streaming-Data

Multithreading Java program to list all frequent path-traversal patterns with increasing size given a minimum support threshold s and a steam of web-click data. Each web server logs all activities into a stream (or a log file in batch mode) in the format of tuples, each containing an integer i (as user ID), and a reference string r (as web URL). A web-click sequence WCS for a user i can be extract from the web-click data as wcsi = [r1, r2, …, rk]. A maximal forward reference MFR is a forward reference path without any backward reference. So, each wcsi can be converted into several MFRs, i.e., wcsi = [mfr1, mfr2, …, mfrj].

Language: Java - Size: 13.7 KB - Last synced at: over 2 years ago - Pushed at: over 4 years ago - Stars: 2 - Forks: 0