I. Muslea, S. Minton, and C. Knoblock. Selective sampling with reduntant views. In Proc. 17th Nat. Conf. Artificial Intelligence, 2000. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Active Learning Selection Strategies for Information Extraction - Finn, Limerick (2003) (Correct)
....members are chosen for labeling. In fact, committee based approaches can be regarded as confidence based, where the confidence in a prediction is based on the disagreement among committee members about that prediction. There has been some work in the application of active learning to IE (eg [13, 11, 12]) 12] use learningalgorithm specific heuristics to choose the next document for annotation. Specifically, their AL algorithm for learning Hidden Markov Models (HMM) identifies difficult unlabeled tokens and asks the user to label them. Difficulty is estimated by the difference between the ....
....most likely and second most likely state of the HMM. Other applications of AL and IE do not rely on a specific learning algorithm. 13] use certainty based sampling, where the certainty of an extracted field is the minimum of the training set accuracies of the rules that extracted the fragment. [11] describe a multi view approach to i.e. Multi view AL a committee based approach in which the committee members are formed by training on different sets of features. Muslea et al. learn two different models for extraction based on two different views of the data, and select the document where both ....
I. Muslea, S. Minton, and C. Knoblock. Selective sampling with reduntant views. In Proc. 17th Nat. Conf. Artificial Intelligence, 2000.
Active Learning Selection Strategies for Information Extraction - Finn, Kushmerick (2003) (Correct)
....and asking the user to label them. Difficulty is estimated by the difference between the most likely and second most likely state of the HMM. Both of these strategies use document selection techniques that are particular to the IE algorithm. A more general multiview strategy is described by [Muslea et al. 2000] and applied to a wrapper induction task. In their case the different views are created using forward and backward rules. ....
I. Muslea, S. Minton, and C. Knoblock. Selective sampling with reduntant views. In Proc. 17th Nat. Conf. Artificial Intelligence, 2000.
Active Learning Selection Strategies for Information Extraction - Aidan Finn Nicholas (Correct)
No context found.
I. Muslea, S. Minton, and C. Knoblock. Selective sampling with reduntant views. In Proc. 17th Nat. Conf. Artificial Intelligence, 2000.
Active Learning Selection Strategies for Information Extraction - Aidan Finn Nicholas (2003) (Correct)
No context found.
I. Muslea, S. Minton, and C. Knoblock. Selective sampling with reduntant views. In Proc. 17th Nat. Conf. Artificial Intelligence, 2000.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC