Prototype-Driven Learning for Sequence Models

@inproceedings{Haghighi2006PrototypeDrivenLF,
  title={Prototype-Driven Learning for Sequence Models},
  author={A. Haghighi and D. Klein},
  booktitle={HLT-NAACL},
  year={2006}
}
  • A. Haghighi, D. Klein
  • Published in HLT-NAACL 2006
  • Computer Science
  • We investigate prototype-driven learning for primarily unsupervised sequence modeling. Prior knowledge is specified declaratively, by providing a few canonical examples of each target annotation label. This sparse prototype information is then propagated across a corpus using distributional similarity features in a log-linear generative model. On part-of-speech induction in English and Chinese, as well as an information extraction task, prototype features provide substantial error rate… CONTINUE READING
    215 Citations
    Controlling Complexity in Part-of-Speech Induction
    • 14
    • PDF
    Semi-supervised SRL System with Bayesian Inference
    • 1
    • PDF
    Semi-Supervised Sequence Labeling with Self-Learned Features
    • 19
    • Highly Influenced
    • PDF
    Syntactic Category Learning as Iterative Prototype-Driven Clustering
    • PDF
    Conditional Random Field Autoencoders for Unsupervised Structured Prediction
    • 68
    • PDF
    Guiding Semi-Supervision with Constraint-Driven Learning
    • 211
    • PDF
    Graphical Models for Primarily Unsupervised Sequence Labeling
    • 1
    • Highly Influenced
    • PDF
    Painless Unsupervised Learning with Features
    • 228
    • PDF
    Weakly-Supervised Learning with Cost-Augmented Contrastive Estimation
    • 5
    • PDF
    Unsupervised Multilingual Learning for POS Tagging
    • 63
    • PDF

    References

    SHOWING 1-10 OF 17 REFERENCES
    Unsupervised Learning of Field Segmentation Models for Information Extraction
    • 88
    • PDF
    Unsupervised induction of stochastic context-free grammars using distributional clustering
    • 124
    • PDF
    Tagging text with a probabilistic model
    • B. Mérialdo
    • Computer Science
    • [Proceedings] ICASSP 91: 1991 International Conference on Acoustics, Speech, and Signal Processing
    • 1991
    • 97
    Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data
    • 11,901
    • Highly Influential
    • PDF
    Morphological features help POS tagging of unknown words across language varieties
    • 82
    • PDF
    Tagging English Text with a Probabilistic Model
    • 619
    • PDF
    Building a Large Annotated Corpus of English: The Penn Treebank
    • 7,633
    • Highly Influential
    • PDF
    Foundations of statistical natural language processing
    • 6,665
    • PDF
    Building a Large-Scale Annotated Chinese Corpus
    • 146
    • PDF