PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Adaptive XML Tree Mining on Evolving Data Streams
Albert Bifet and Ricard Gavaldà
In: 7th International Workshop on Mining and Learning with Graphs MLG 2009, July 2-4, 2009, Leuven, Belgium.


We propose a new method to classify trees, using closed and maximal frequent trees. Closed trees maintain the same information as frequent trees using less space and maximal trees maintain approximate information. We use them to reduce the number of classification features. We present a new framework for data stream tree classification. For the first component of our classification framework, using a methodology based in Galois Lattice Theory, we present three closed tree mining algorithms: an incremental one IncTreeMiner, a sliding-window based one, WinTreeMiner, and finally one that mines closed trees adaptively from data streams, AdaTreeMiner. To the best of our knowledge this is the first work on tree classification in streaming data varying with time. We give a first experimental evaluation of the proposed classification method.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Poster)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
Theory & Algorithms
ID Code:5625
Deposited By:Albert Bifet
Deposited On:08 March 2010