Overall Objectives
Research Program
Application Domains
Highlights of the Year
New Software and Platforms
New Results
Bilateral Contracts and Grants with Industry
Partnerships and Cooperations
XML PDF e-pub
PDF e-Pub

Section: New Results

Negative Temporal Sequence Mining

Participants : Katerina Tsesmeli, Thomas Guyet, René Quiniou, Manel Boumghar [EDF R&D] , Laurent Pierre [EDF R&D] .

Temporal pattern mining is one of the important tasks in the data mining research field. It aims at extracting interesting sequences of occurring events from timestamped event sequences as well as their temporal constraints relating sequence events. Little research has focused on mining sequential patterns with non-occurring (negative) events, though they can bring much value and relevance to extracted patterns. In this context, we are interested in formalizing normal and undesirable situations, that can be defined in terms of negative temporal patterns. We proposed the NTGSP algorithm [17] that extracts frequent sequences with positive and negative events, as well as temporal information about the delay between these events. The method performance has been evaluated on synthetic sequences and on commercial data provided by EDF , a major french power distribution company.