Ensemble of classifiers for noise detection in pos tagged corpora

Abstract

In this paper we apply the ensemble approach to the identification of incorrectly annotated items (noise) in a training set. In a controlled experiment, memory-based, decision tree-based and transformation-based classifiers are used as a filter to detect and remove noise deliberately introduced into a manually tagged corpus. The results indicate that the method can be successfully applied to automatically detect errors in a corpus.