Computer Science – Computation and Language
Scientific paper
2001-07-17
Proceedings of the 5th Computational Natural Language Learning Workshop (CoNNL-2001), pages 1-8, Toulouse, France
Computer Science
Computation and Language
8 pages, 2 figures, presented at CONLL 2001
Scientific paper
This paper presents a novel method that allows a machine learning algorithm following the transformation-based learning paradigm \cite{brill95:tagging} to be applied to multiple classification tasks by training jointly and simultaneously on all fields. The motivation for constructing such a system stems from the observation that many tasks in natural language processing are naturally composed of multiple subtasks which need to be resolved simultaneously; also tasks usually learned in isolation can possibly benefit from being learned in a joint framework, as the signals for the extra tasks usually constitute inductive bias. The proposed algorithm is evaluated in two experiments: in one, the system is used to jointly predict the part-of-speech and text chunks/baseNP chunks of an English corpus; and in the second it is used to learn the joint prediction of word segment boundaries and part-of-speech tagging for Chinese. The results show that the simultaneous learning of multiple tasks does achieve an improvement in each task upon training the same tasks sequentially. The part-of-speech tagging result of 96.63% is state-of-the-art for individual systems on the particular train/test split.
Florian Radu
Ngai Grace
No associations
LandOfFree
Multidimensional Transformation-Based Learning does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Multidimensional Transformation-Based Learning, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multidimensional Transformation-Based Learning will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-642654