Discovering Fuzzy Unexpected Sequences with Beliefs

Haoyuan Li (1, 2) , Anne Laurent (3) , Pascal Poncelet (3)

1 LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
2 LGI2P - Laboratoire de Génie Informatique et Ingénierie de Production
3 TATOO - Fouille de données environnementales

In this paper we present a novel approach for discovering fuzzy unexpected sequences, such as certainly unexpected, almost unexpected and a little unexpected, from databases with respect to user defined beliefs. We first formalize the belief on sequences and the different types of unexpectedness, then we detail the algorithm Taufu that finds fuzzy unexpected sequences with beliefs. Our approach has been verified with various experiments.

Domaines: Base de données [cs.DB]

Haoyuan Li, Anne Laurent, Pascal Poncelet. Discovering Fuzzy Unexpected Sequences with Beliefs. IPMU'08: 12th International Conference of Information Processing and Management of Uncertainty in Knowledge-Based Systems, pp.1709-1716. ⟨lirmm-00275951⟩