Named entity recognition through classifier combination
Radu Florian, Abe Ittycheriah, Hongyan Jing et al. · 2003
This paper presents a classifier-combination experimental framework for named entity recognition in which four diverse classifiers (robust linear classifier, maximum entropy, transformation-based learning, and hidden Markov model) are combined under different conditions. When no gazetteer or other additional training resources are used, the combined system attains a performance of 91.6F on the English development data; integrating name, location and person gazetteers, and named entity systems trained on additional, more general, data reduces the F-measure error by a factor of 15 to 21% on the English data.
Cited by 399 works. doi:10.3115/1119176.1119201 Open-access PDF
Authors
References
- Error bounds for convolutional codes and an asymptotically optimum decoding algorithm
- A maximum entropy approach to natural language processing
- Statistical Methods for Speech Recognition
- Generalized Iterative Scaling for Log-Linear Models
- Transformation-based error-driven learning and natural language processing: a case study in part-of-speech tagging
- An Algorithm that Learns What's in a Name
- A maximum entropy approach to named entity recognition
- A robust risk minimization based named entity recognition system
- Text chunking based on a generalization of winnow
- IBM's Statistical Question Answering System.