Combining labeled and unlabeled data with co-training
COLT' 98 Proceedings of the eleventh annual conference on Computational learning theory
Foundations of statistical natural language processing
Foundations of statistical natural language processing
An Iterative Approach to Word Sense Disambiguation
Proceedings of the Thirteenth International Florida Artificial Intelligence Research Society Conference
Natural Language Engineering
Unsupervised word sense disambiguation rivaling supervised methods
ACL '95 Proceedings of the 33rd annual meeting on Association for Computational Linguistics
A method for word sense disambiguation of unrestricted text
ACL '99 Proceedings of the 37th annual meeting of the Association for Computational Linguistics on Computational Linguistics
| Hi-index | 0.00 |
The task of disambiguation is to determine which of the senses of an ambiguous word is invoked in a particular use of the word [5,8]. It is known that the statistical methods produce high accuracy results for semantically tagged corpora [2]. Also, Word Net is a good source of information for WSD [3,4]. Since for Romanian language does not exist neither a corpus nor something similar with WordNet, we propose an algorithm for WSD which requires only information that can be extracted from untagged corpus. Our algorithm preserves the advantage of principles of Yarowsky [9,7,10] and adds the known high performance of a NBC algorithms. It learns to make predictions based on local context with only a few labeled contexts and many unlabeled ones.