A framework for multiple-instance learning
NIPS '97 Proceedings of the 1997 conference on Advances in neural information processing systems 10
EMNLP '02 Proceedings of the ACL-02 conference on Empirical methods in natural language processing - Volume 10
Online Passive-Aggressive Algorithms
The Journal of Machine Learning Research
Exploring hedge identification in biomedical literature
Journal of Biomedical Informatics
Learning the scope of hedge cues in biomedical texts
BioNLP '09 Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing
The CoNLL-2010 shared task: learning to detect hedges and their scope in natural language text
CoNLL '10: Shared Task Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task
A cascade method for detecting hedges and their scope in natural language text
CoNLL '10: Shared Task Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task
A hedgehop over a max-margin framework using hedge cues
CoNLL '10: Shared Task Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task
Detecting hedge cues and their scopes with average perceptron
CoNLL '10: Shared Task Proceedings of the Fourteenth Conference on Computational Natural Language Learning --- Shared Task
Hi-index | 0.00 |
Distinguishing uncertain information from factual ones in online texts is of essential importance in information extraction, because uncertain information would mislead systems to find useless even fault information. In this paper, we propose a method for detecting uncertain sentences with multiple instance learning (MIL). Based on the basic assumption, we derive two new constraints for estimating the weight vector by defining a probability margin, which is used in an online learning algorithm known as Passive-Aggressive algorithm. To demonstrate the effectiveness of our method, we experiment on the biomedical corpus. Compared with an intuitive method with conventional single instance learning (SIL), our method provide higher performance by raising the performance from 79.07% up to 82.55%, over 3% improvement.