Dutch Named Entity Recognition using Classifier Ensembles
This paper explores the use of classifier ensembles for the task of named entity recognition (NER) on a Dutch dataset. Classifiers from 3 classification frameworks, namely memory-based learning (MBL), conditional random fields (CRF) and support vector machines (SVM), were trained on 8 different feature sets to create a pool of classifiers from which an ensemble could be built. A genetic algorithm approach was used to find the optimal ensemble combination, given various voting mechanisms for combining classifier outputs. The experiments yielded a classifier ensemble that outperformed the best i... Mehr ...
Verfasser: | |
---|---|
Dokumenttyp: | Part of book or chapter of book |
Erscheinungsdatum: | 2010 |
Schlagwörter: | Taalwetenschap |
Sprache: | Englisch |
Permalink: | https://search.fid-benelux.de/Record/base-29038114 |
Datenquelle: | BASE; Originalkatalog |
Powered By: | BASE |
Link(s) : | https://dspace.library.uu.nl/handle/1874/297149 |
This paper explores the use of classifier ensembles for the task of named entity recognition (NER) on a Dutch dataset. Classifiers from 3 classification frameworks, namely memory-based learning (MBL), conditional random fields (CRF) and support vector machines (SVM), were trained on 8 different feature sets to create a pool of classifiers from which an ensemble could be built. A genetic algorithm approach was used to find the optimal ensemble combination, given various voting mechanisms for combining classifier outputs. The experiments yielded a classifier ensemble that outperformed the best individual classifier by 0.67 percentage points (F-score), a small but statistically significant margin. Experimental results also showed that ensembling classifiers from different frameworks benefits generalization performance.