Detecting annotation noise in automatically labelled data
We introduce a method for error detection in automatically annotated text, aimed at supporting the creation of high-quality language resources at affordable cost. Our method combines an unsupervised generative model with human supervision from active learning. We test our approach on in-domain and o...
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Dokumenttyp: | Kapitel/Artikel Konferenzschrift |
| Sprache: | Englisch |
| Veröffentlicht: |
July 2017
|
| In: |
The 55th Annual Meeting of the Association for Computational Linguistics - proceedings of the conference ; Vol. 1: Long papers
Year: 2017, Pages: 1160-1170 |
| DOI: | 10.18653/v1/P17-1107 |
| Online-Zugang: | Verlag, lizenzpflichtig, Volltext: https://doi.org/10.18653/v1/P17-1107 Verlag, lizenzpflichtig, Volltext: https://www.aclweb.org/anthology/P17-1107 
|
| Verfasserangaben: | Ines Rehbein, Josef Ruppenhofer |
| Zusammenfassung: | We introduce a method for error detection in automatically annotated text, aimed at supporting the creation of high-quality language resources at affordable cost. Our method combines an unsupervised generative model with human supervision from active learning. We test our approach on in-domain and out-of-domain data in two languages, in AL simulations and in a real world setting. For all settings, the results show that our method is able to detect annotation errors with high precision and high recall. |
|---|---|
| Beschreibung: | Gesehen am 31.03.2020 |
| Beschreibung: | Online Resource |
| ISBN: | 9781945626753 |
| DOI: | 10.18653/v1/P17-1107 |