TY - GEN
T1 - Structured output learning with indirect supervision
AU - Chang, Ming Wei
AU - Srikumar, Vivek
AU - Goldwasser, Dan
AU - Roth, Dan
PY - 2010/9/17
Y1 - 2010/9/17
N2 - We present a novel approach for structure prediction that addresses the difficulty of obtaining labeled structures for training. We observe that structured output problems often have a companion learning problem of determining whether a given input possesses a good structure. For example, the companion problem for the part-of-speech (POS) tagging task asks whether a given sequence of words has a corresponding sequence of POS tags that is "legitimate". While obtaining direct supervision for structures is difficult and expensive, it is often very easy to obtain indirect supervision from the companion binary decision problem. In this paper, we develop a large margin framework that jointly learns from both direct and indirect forms of supervision. Our experiments ex-hibit the significant contribution of the easy-to-get indirect binary supervision on three important NLP structure learning problems.
AB - We present a novel approach for structure prediction that addresses the difficulty of obtaining labeled structures for training. We observe that structured output problems often have a companion learning problem of determining whether a given input possesses a good structure. For example, the companion problem for the part-of-speech (POS) tagging task asks whether a given sequence of words has a corresponding sequence of POS tags that is "legitimate". While obtaining direct supervision for structures is difficult and expensive, it is often very easy to obtain indirect supervision from the companion binary decision problem. In this paper, we develop a large margin framework that jointly learns from both direct and indirect forms of supervision. Our experiments ex-hibit the significant contribution of the easy-to-get indirect binary supervision on three important NLP structure learning problems.
UR - http://www.scopus.com/inward/record.url?scp=77956553315&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77956553315&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:77956553315
SN - 9781605589077
T3 - ICML 2010 - Proceedings, 27th International Conference on Machine Learning
SP - 199
EP - 206
BT - ICML 2010 - Proceedings, 27th International Conference on Machine Learning
T2 - 27th International Conference on Machine Learning, ICML 2010
Y2 - 21 June 2010 through 25 June 2010
ER -