# Chordalysis: Free software for Log-linear analysis of Big Data

Chordalysis is a log-linear analysis method for big data, which exploits recent discoveries in graph theory by representing complex models as compositions of triangular structures (aka chordal graphs).

**By Francois Petitjean, Monash University.**

Log-linear analysis is the statistical method used to capture multi-way relationships between variables. However, due to its exponential nature, previous approaches did not allow scale-up to more than a dozen variables.

Chordalysis is a method that scales log-linear analysis to high-dimensional data (which is, in our case, identical to learning the structure of Bayesian Networks and/or Markov Networks). Chordalysis exploits recent discoveries in graph theory by representing complex models as compositions of triangular structures, also known as chordal graphs.

Chordalysis is the software resulting from years of research in graphical modeling; associated scientific publications are:

- KDD 2016: A multiple test correction for streams and cascades of statistical hypothesis tests
- SDM 2015: Scaling log-linear analysis to datasets with thousands of variables
- ICDM 2014: A statistically efficient and scalable method for log-linear analysis of high-dimensional data
- ICDM 2013: Scaling log-linear analysis to high-dimensional data

Chordalysis software is released under GPL and is available at:

github.com/fpetitjean/Chordalysis

It also has an R interface.

More information on François' website