Weakly Supervised Natural Language Learning Without Redundant Views

Vincent Ng and Claire Cardie.
HLT-NAACL 2003: Proceedings of the Main Conference, pp. 173-180, 2003.

Click here for the PostScript or PDF version. The talk slides are available here.

Abstract

We investigate single-view algorithms as an alternative to multi-view algorithms for weakly supervised learning for natural language processing tasks without a natural feature split. In particular, we apply co-training, self-training, and EM to one such task and find that both self-training and FS-EM, a new variation of EM that incorporates feature selection, outperform co-training and are comparatively less sensitive to parameter changes.

BibTeX entry

@InProceedings{Ng+Cardie:03a,
  author = {Vincent Ng and Claire Cardie},
  title = {Weakly Supervised Natural Language Learning Without Redundant Views},
  booktitle = {HLT-NAACL 2003: Proceedings of the Main Conference},
  pages = {173--180},
  year = 2003
}