Resolving "orphaned" non-specific structures using machine learning and natural language processing methods

Dongfang Xu, Steven S. Chong, Thomas Rodenhausen, Hong Cui

Research output: Contribution to journalArticle

Abstract

Scholarly publications of biodiversity literature contain a vast amount of information in human readable format. The detailed morphological descriptions in these publications contain rich information that can be extracted to facilitate analysis and computational biology research. However, the idiosyncrasies of morphological descriptions still pose a number of challenges to machines. In this work, we demonstrate the use of two different approaches to resolve meronym (i.e. part-of) relations between anatomical parts and their anchor organs, including a syntactic rule-based approach and a SVM-based (support vector machine) method. Both methods made use of domain ontologies. We compared the two approaches with two other baseline methods and the evaluation results show the syntactic methods (92.1% F1 score) outperformed the SVM methods (80.7% F1 score) and the part-of ontologies were valuable knowledge sources for the task. It is notable that the mistakes made by the two approaches rarely overlapped. Additional tests will be conducted on the development version of the Explorer of Taxon Concepts toolkit before we make the functionality publicly available. Meanwhile, we will further investigate and leverage the complementary nature of the two approaches to further drive down the error rate, as in practical application, even a 1% error rate could lead to hundreds of errors.

Original languageEnglish (US)
Article numbere26659
JournalBiodiversity Data Journal
Volume6
DOIs
StatePublished - Jan 1 2018

Fingerprint

artificial intelligence
processing technology
methodology
anchor
bioinformatics
machine learning
method
biodiversity
testing
rate

Keywords

  • Anaphora resolution
  • Biodiversity literature
  • Information extraction
  • Machine learning
  • Morphological descriptions
  • Ontology application
  • Performance evaluation

ASJC Scopus subject areas

  • Ecology, Evolution, Behavior and Systematics
  • Ecology

Cite this

Resolving "orphaned" non-specific structures using machine learning and natural language processing methods. / Xu, Dongfang; Chong, Steven S.; Rodenhausen, Thomas; Cui, Hong.

In: Biodiversity Data Journal, Vol. 6, e26659, 01.01.2018.

Research output: Contribution to journalArticle

@article{8d3666f15d7d463281327de300bf8afd,
title = "Resolving {"}orphaned{"} non-specific structures using machine learning and natural language processing methods",
abstract = "Scholarly publications of biodiversity literature contain a vast amount of information in human readable format. The detailed morphological descriptions in these publications contain rich information that can be extracted to facilitate analysis and computational biology research. However, the idiosyncrasies of morphological descriptions still pose a number of challenges to machines. In this work, we demonstrate the use of two different approaches to resolve meronym (i.e. part-of) relations between anatomical parts and their anchor organs, including a syntactic rule-based approach and a SVM-based (support vector machine) method. Both methods made use of domain ontologies. We compared the two approaches with two other baseline methods and the evaluation results show the syntactic methods (92.1{\%} F1 score) outperformed the SVM methods (80.7{\%} F1 score) and the part-of ontologies were valuable knowledge sources for the task. It is notable that the mistakes made by the two approaches rarely overlapped. Additional tests will be conducted on the development version of the Explorer of Taxon Concepts toolkit before we make the functionality publicly available. Meanwhile, we will further investigate and leverage the complementary nature of the two approaches to further drive down the error rate, as in practical application, even a 1{\%} error rate could lead to hundreds of errors.",
keywords = "Anaphora resolution, Biodiversity literature, Information extraction, Machine learning, Morphological descriptions, Ontology application, Performance evaluation",
author = "Dongfang Xu and Chong, {Steven S.} and Thomas Rodenhausen and Hong Cui",
year = "2018",
month = "1",
day = "1",
doi = "10.3897/BDJ.6.e26659",
language = "English (US)",
volume = "6",
journal = "Biodiversity Data Journal",
issn = "1314-2828",
publisher = "Pensoft Publishers",

}

TY - JOUR

T1 - Resolving "orphaned" non-specific structures using machine learning and natural language processing methods

AU - Xu, Dongfang

AU - Chong, Steven S.

AU - Rodenhausen, Thomas

AU - Cui, Hong

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Scholarly publications of biodiversity literature contain a vast amount of information in human readable format. The detailed morphological descriptions in these publications contain rich information that can be extracted to facilitate analysis and computational biology research. However, the idiosyncrasies of morphological descriptions still pose a number of challenges to machines. In this work, we demonstrate the use of two different approaches to resolve meronym (i.e. part-of) relations between anatomical parts and their anchor organs, including a syntactic rule-based approach and a SVM-based (support vector machine) method. Both methods made use of domain ontologies. We compared the two approaches with two other baseline methods and the evaluation results show the syntactic methods (92.1% F1 score) outperformed the SVM methods (80.7% F1 score) and the part-of ontologies were valuable knowledge sources for the task. It is notable that the mistakes made by the two approaches rarely overlapped. Additional tests will be conducted on the development version of the Explorer of Taxon Concepts toolkit before we make the functionality publicly available. Meanwhile, we will further investigate and leverage the complementary nature of the two approaches to further drive down the error rate, as in practical application, even a 1% error rate could lead to hundreds of errors.

AB - Scholarly publications of biodiversity literature contain a vast amount of information in human readable format. The detailed morphological descriptions in these publications contain rich information that can be extracted to facilitate analysis and computational biology research. However, the idiosyncrasies of morphological descriptions still pose a number of challenges to machines. In this work, we demonstrate the use of two different approaches to resolve meronym (i.e. part-of) relations between anatomical parts and their anchor organs, including a syntactic rule-based approach and a SVM-based (support vector machine) method. Both methods made use of domain ontologies. We compared the two approaches with two other baseline methods and the evaluation results show the syntactic methods (92.1% F1 score) outperformed the SVM methods (80.7% F1 score) and the part-of ontologies were valuable knowledge sources for the task. It is notable that the mistakes made by the two approaches rarely overlapped. Additional tests will be conducted on the development version of the Explorer of Taxon Concepts toolkit before we make the functionality publicly available. Meanwhile, we will further investigate and leverage the complementary nature of the two approaches to further drive down the error rate, as in practical application, even a 1% error rate could lead to hundreds of errors.

KW - Anaphora resolution

KW - Biodiversity literature

KW - Information extraction

KW - Machine learning

KW - Morphological descriptions

KW - Ontology application

KW - Performance evaluation

UR - http://www.scopus.com/inward/record.url?scp=85057456634&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057456634&partnerID=8YFLogxK

U2 - 10.3897/BDJ.6.e26659

DO - 10.3897/BDJ.6.e26659

M3 - Article

AN - SCOPUS:85057456634

VL - 6

JO - Biodiversity Data Journal

JF - Biodiversity Data Journal

SN - 1314-2828

M1 - e26659

ER -