Filling the gap: semi-supervised learning for opinion detection across domains
CoNLL '11 Proceedings of the Fifteenth Conference on Computational Natural Language Learning
| Hi-index | 0.00 |
Research on opinion detection has shown that a large number of opinion-labeled data are necessary for capturing subtle opinions. However, opinion-labeled data, especially at the sub-document level, are often limited. This paper describes the application of Semi-Supervised Learning (SSL) to automatically produce more labeled data and explores the potential of SSL to improve transfer of labeled data to new domains. Preliminary results show that SSL performance is very close to a supervised system trained on the full data set and improves performance on out-of-domain data.