PhD. candidate, advised by
Sihem Amer-Yahia &
Funded by the
Member of the SLIDE team.
Last updated: September, 27th 2016.
I am adapting a selection of pattern enumeration algorithms to big data:
such data mining techniques should remain relevant despite the
long-tail distribution of the content,
and should leverage parallel and distributed systems.
Mining and ranking closed itemsets from large-scale transactional datasets, PhD thesis
Testing Interestingness Measures in Practice: A Large-Scale Analysis of Buying Patterns,
with Martin Kirchgessner, Vincent Leroy, Sihem Amer-Yahia, Shashwat Mishra and Intermarché Alimentaire International.
To be presented at the 3rd IEEE International Conference on Data Science and Advanced Analytics (IEEE DSAA 2016), in Montreal, Canada, on October 17-19 2016.
TopPI: An Efficient Algorithm for Item-Centric Mining,
with Vincent Leroy, Alexandre Termier, Sihem Amer-Yahia and Marie-Christine Rousset,
in Proceedings of the 18th International Conference on Big Data Analytics and Knowledge Discovery, DaWaK 2016, presented in Porto, Portugal on September 6th, 2016.
G5K-v11n, a metrics visualization toolkit for the
a multi-threaded implementation of the LCM (Linear Closed itemsets Miner)
algorithm, in Java, as proposed by T.Uno & H.Arimura.
It is available as a library through Maven's central repositories.
Pattern mining in big datasets corners,
a poster presented at the french conference BDA 2013
during the "young researchers" session.
Born in 1987, Martin firstly encountered a command line around 1995 and
wrote his first lines of code (Visual Basic !) in 2000. He graduated
from Grenoble INP's Ensimag in 2008 (Master degree in computer engineering).
After three years spent in an entrepreneurship adventure,
he worked as a Web application developer
and then considered starting a PhD.
He joined the LIG in 2012, starting its pattern mining research with
Marie-Christine Rousset and
By the way: Kirchgessner
is (simply) pronounced as if it was written "Kirgessner".