Exploring a Multi-Layered Cross-Genre Corpus of Document-Level Semantic Relations

Author:

Williamson Gregor1,Cao Angela1,Chen Yingying1,Ji Yuxin1,Xu Liyan1,Choi Jinho D.1ORCID

Affiliation:

1. Department of Computer Science, Emory University, Atlanta, GA 30322, USA

Abstract

This paper introduces a multi-layered cross-genre corpus, annotated for coreference resolution, causal relations, and temporal relations, comprising a variety of genres, from news articles and children’s stories to Reddit posts. Our results reveal distinctive genre-specific characteristics at each layer of annotation, highlighting unique challenges for both annotators and machine learning models. Children’s stories feature linear temporal structures and clear causal relations. In contrast, news articles employ non-linear temporal sequences with minimal use of explicit causal or conditional language and few first-person pronouns. Lastly, Reddit posts are author-centered explanations of ongoing situations, with occasional meta-textual reference. Our annotation schemes are adapted from existing work to better suit a broader range of text types. We argue that our multi-layered cross-genre corpus not only reveals genre-specific semantic characteristics but also indicates a rich contextual interplay between the various layers of semantic information. Our MLCG corpus is shared under the open-source Apache 2.0 license.

Funder

Amazon Alexa AI

Publisher

MDPI AG

Subject

Information Systems

Reference45 articles.

1. Li, X., Palmer, M., Xue, N., Ramshaw, L., Maamouri, M., Bies, A., Conger, K., Grimes, S., and Strassel, S. (2016, January 23–28). Large Multi-lingual, Multi-level and Multi-genre Annotation Corpus. Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16), Portorož, Slovenia.

2. Zeldes, A., and Simonson, D. (2016, January 11). Different Flavors of GUM: Evaluating Genre and Sentence Type Effects on Multilayer Corpus Annotation Quality. Proceedings of the 10th Linguistic Annotation Workshop Held in Conjunction with ACL 2016 (LAW-X 2016), Berlin, Germany.

3. The GUM corpus: Creating multilayer resources in the classroom;Zeldes;Lang. Resour. Eval.,2017

4. Zeldes, A. (2018). Multilayer Corpus Studies, Routledge.

5. Gessler, L., Peng, S., Liu, Y., Zhu, Y., Behzad, S., and Zeldes, A. (2020, January 11–16). AMALGUM—A Free, Balanced, Multilayer English Web Corpus. Proceedings of the 12th Language Resources and Evaluation Conference, Marseille, France.

Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3