Process Mining and big data: what service and library creators don't tell you
summary
The talk examines practical engineering challenges of applying Process Mining to large-scale event data, showing that popular libraries degrade on big datasets. It analyzes tools such as pm4py, retentioneering and SberGraph, compares their pros and cons, and presents a pipeline of optimization techniques and ready-made approaches for speeding up graph construction and processing.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 3:15Business problem and metrics
- 9:54Why basic analytics fails
- 14:16What is process mining
- 19:15Event Log as foundation
- 23:29Data collection infrastructure
- 30:25Process mining solutions
- 35:17Benchmark of solutions
- 37:02Adapting process mining
- 43:01Summary
related talks
34mProcess mining from the inside
24mScalable Stream Processing and Large-Scale Data Analysis
44mProcess and analyze millions of events per second with open source tools
45mBig Data — How to Mine Gold from Data
19mDemystifying Big Data
07mMining the Field of Big Data Strategies
19mAdoption of Big Data in Securities Processing Industry A Developer's Perspective
41mBig Data and Ruby