C. Scheible, R. Klinger, und S. Padó. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Seite 1736--1745. Berlin, Germany, Association for Computational Linguistics, (August 2016)
DOI: 10.18653/v1/P16-1164
Zusammenfassung
Quotation detection is the task of locating spans of quoted speech in text. The state of the art treats this problem as a sequence labeling task and employs linear-chain conditional random fields. We question the efficacy of this choice: The Markov assumption in the model prohibits it from making joint decisions about the begin, end, and internal context of a quotation. We perform an extensive analysis with two new model architectures. We find that (a), simple boundary classification combined with a greedy prediction strategy is competitive with the state of the art; (b), a semi-Markov model significantly outperforms all others, by relaxing the Markov assumption.
Beschreibung
Model Architectures for Quotation Detection - ACL Anthology
%0 Conference Paper
%1 scheible2016model
%A Scheible, Christian
%A Klinger, Roman
%A Padó, Sebastian
%B Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%C Berlin, Germany
%D 2016
%I Association for Computational Linguistics
%K citation detection extraction language natural nlp processing quotation quote
%P 1736--1745
%R 10.18653/v1/P16-1164
%T Model Architectures for Quotation Detection
%U https://www.aclweb.org/anthology/P16-1164
%X Quotation detection is the task of locating spans of quoted speech in text. The state of the art treats this problem as a sequence labeling task and employs linear-chain conditional random fields. We question the efficacy of this choice: The Markov assumption in the model prohibits it from making joint decisions about the begin, end, and internal context of a quotation. We perform an extensive analysis with two new model architectures. We find that (a), simple boundary classification combined with a greedy prediction strategy is competitive with the state of the art; (b), a semi-Markov model significantly outperforms all others, by relaxing the Markov assumption.
@inproceedings{scheible2016model,
abstract = {Quotation detection is the task of locating spans of quoted speech in text. The state of the art treats this problem as a sequence labeling task and employs linear-chain conditional random fields. We question the efficacy of this choice: The Markov assumption in the model prohibits it from making joint decisions about the begin, end, and internal context of a quotation. We perform an extensive analysis with two new model architectures. We find that (a), simple boundary classification combined with a greedy prediction strategy is competitive with the state of the art; (b), a semi-Markov model significantly outperforms all others, by relaxing the Markov assumption.},
added-at = {2020-10-20T11:22:35.000+0200},
address = {Berlin, Germany},
author = {Scheible, Christian and Klinger, Roman and Padó, Sebastian},
biburl = {https://www.bibsonomy.org/bibtex/216e86e0db4ede7b6a1a29e24aee776e5/jaeschke},
booktitle = {Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
description = {Model Architectures for Quotation Detection - ACL Anthology},
doi = {10.18653/v1/P16-1164},
interhash = {c77dfb02001fe26838c9936221ace71a},
intrahash = {16e86e0db4ede7b6a1a29e24aee776e5},
keywords = {citation detection extraction language natural nlp processing quotation quote},
month = aug,
pages = {1736--1745},
publisher = {Association for Computational Linguistics},
timestamp = {2020-10-20T11:22:35.000+0200},
title = {Model Architectures for Quotation Detection},
url = {https://www.aclweb.org/anthology/P16-1164},
year = 2016
}