DMCA
Exploiting feature hierarchy for transfer learning in named entity recognition (2008)
Cached
Download Links
| Venue: | In ACL:HLT ’08 |
| Citations: | 13 - 3 self |
BibTeX
@INPROCEEDINGS{Arnold08exploitingfeature,
author = {Andrew Arnold and Ramesh Nallapati and William W. Cohen},
title = {Exploiting feature hierarchy for transfer learning in named entity recognition},
booktitle = {In ACL:HLT ’08},
year = {2008}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We present a novel hierarchical prior structure for supervised transfer learning in named entity recognition, motivated by the common structure of feature spaces for this task across natural language data sets. The problem of transfer learning, where information gained in one learning task is used to improve performance in another related task, is an important new area of research. In the subproblem of domain adaptation, a model trained over a source domain is generalized to perform well on a related target domain, where the two domains’ data are distributed similarly, but not identically. We introduce the concept of groups of closely-related domains, called genres, and show how inter-genre adaptation is related to domain adaptation. We also examine multitask learning, where two domains may be related, but where the concept to be learned in each case is distinct. We show that our prior conveys useful information across domains, genres and tasks, while remaining robust to spurious signals not related to the target domain and concept. We further show that our model generalizes a class of similar hierarchical priors, smoothed to varying degrees, and lay the groundwork for future exploration in this area. 1
Keyphrases
entity recognition transfer learning feature hierarchy multitask learning feature space supervised transfer learning target domain related task natural language data set similar hierarchical prior source domain domain adaptation novel hierarchical prior structure inter-genre adaptation spurious signal prior conveys useful information domain data common structure future exploration learning task related target domain important new area closely-related domain
