Non-lexical neural architecture for fine-grained POS Tagging

Matthieu Labeau, Kevin Löser, Alexandre Allauzen

Abstract

In this paper we explore a POS tagging application of neural architectures that can infer word representations from the raw character stream. It relies on two modelling stages that are jointly learnt: a convolutional network that infers a word representation directly from the character stream, followed by a prediction stage. Models are evaluated on a POS and morphological tagging task for German. Experimental results show that the convolutional network can infer meaningful word representations, while for the prediction stage, a well designed and structured strategy allows the model to outperform stateof-the-art results, without any feature engineering.
BibTeX
@inproceedings{labeau2015non,
  author = {Matthieu Labeau and Kevin Löser and Alexandre Allauzen},
  title = {{Non-lexical neural architecture for fine-grained POS Tagging}},
  booktitle = {Conference on Empirical Methods in Natural Language Processing},
  year = {2015},
  doi = {10.18653/v1/d15-1025},
}
Figure