he Possibilities of Automatic Detection/Correction of Errors in Tagged Corpora: A Pilot Study on a German Corpus

Abstract

The performance of taggers is usuallye valuated bytheir percentual success rate. Because of the pure quantitativity of such an approach, all errors committed bythe tagger are treated on a par for the purpose of the evaluation. This paper takes a different, qualitative stand on the topic, arguing that the previous viewpoint is not linguisticallyadequate: the errors (might) differ in severity. General implications for tagging are discussed, and a simple method is proposed and exemplified, able to

• 1. detect and in some cases even rectifythe most severe errors and thus

• 2. contribute to arriving finally at a better tagged corpus.

Some encouraging results achieved bya verysimple, manuallyperformed test and evaluation on a small sample of a corpus are given.

This work has been sponsored bythe Fonds zur Förderung der wissenschaftlichen Forschung (FWF), Grant No. P13224.

The Austrian Research Institute for Artificial Intelligence is supported bythe Austrian Federal Ministry of Education, Science and Culture.