Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods

Stefania Rubrichi, Matteo Gabetta, Riccardo Bellazzi, Cristiana Larizza, Silvana Quaglini

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Information about medications is critical in improving the patients' safety and quality of care. Most adverse drug events are predictable from the known pharmacology of the drugs and many represent known interactions and are, therefore, likely to be preventable. However, most of this information is locked in free-text and, as such, cannot be actively accessed and elaborated by computerized applications. In this work, we propose three different approaches to the problem of automatic recognition of drug-drug interactions that we have developed within the "First Challenge Task: Drug-Drug Interaction Extraction" competition. Our approaches learn to discriminate between semantically interesting and uninteresting content in a structured prediction framework as well as a rule-based one. The systems are trained using the DrugDDI corpus provided by the challenge organizers. An empirical analysis of the three approaches on this dataset shows that the inclusion of rule-based methods is indeed advantageous.

Original languageEnglish
Title of host publicationCEUR Workshop Proceedings
Pages67-74
Number of pages8
Volume761
Publication statusPublished - 2011
Event1st Challenge Task on Drug-Drug Interaction Extraction 2011, DDIExtraction 2011 - Co-located with the 27th Conference of the Spanish Society for Natural Language Processing, SEPLN 2011 - Huelva, Spain
Duration: Sep 7 2011Sep 7 2011

Other

Other1st Challenge Task on Drug-Drug Interaction Extraction 2011, DDIExtraction 2011 - Co-located with the 27th Conference of the Spanish Society for Natural Language Processing, SEPLN 2011
CountrySpain
CityHuelva
Period9/7/119/7/11

Fingerprint

Drug interactions

Keywords

  • Adverse drug events
  • Conditional random fields
  • Drug-Drug Interactions
  • Information extraction
  • Support vector machines

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Rubrichi, S., Gabetta, M., Bellazzi, R., Larizza, C., & Quaglini, S. (2011). Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods. In CEUR Workshop Proceedings (Vol. 761, pp. 67-74)

Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods. / Rubrichi, Stefania; Gabetta, Matteo; Bellazzi, Riccardo; Larizza, Cristiana; Quaglini, Silvana.

CEUR Workshop Proceedings. Vol. 761 2011. p. 67-74.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Rubrichi, S, Gabetta, M, Bellazzi, R, Larizza, C & Quaglini, S 2011, Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods. in CEUR Workshop Proceedings. vol. 761, pp. 67-74, 1st Challenge Task on Drug-Drug Interaction Extraction 2011, DDIExtraction 2011 - Co-located with the 27th Conference of the Spanish Society for Natural Language Processing, SEPLN 2011, Huelva, Spain, 9/7/11.
Rubrichi S, Gabetta M, Bellazzi R, Larizza C, Quaglini S. Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods. In CEUR Workshop Proceedings. Vol. 761. 2011. p. 67-74
Rubrichi, Stefania ; Gabetta, Matteo ; Bellazzi, Riccardo ; Larizza, Cristiana ; Quaglini, Silvana. / Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods. CEUR Workshop Proceedings. Vol. 761 2011. pp. 67-74
@inproceedings{e5da5afdb61b41ca8635338948f5d3b3,
title = "Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods",
abstract = "Information about medications is critical in improving the patients' safety and quality of care. Most adverse drug events are predictable from the known pharmacology of the drugs and many represent known interactions and are, therefore, likely to be preventable. However, most of this information is locked in free-text and, as such, cannot be actively accessed and elaborated by computerized applications. In this work, we propose three different approaches to the problem of automatic recognition of drug-drug interactions that we have developed within the {"}First Challenge Task: Drug-Drug Interaction Extraction{"} competition. Our approaches learn to discriminate between semantically interesting and uninteresting content in a structured prediction framework as well as a rule-based one. The systems are trained using the DrugDDI corpus provided by the challenge organizers. An empirical analysis of the three approaches on this dataset shows that the inclusion of rule-based methods is indeed advantageous.",
keywords = "Adverse drug events, Conditional random fields, Drug-Drug Interactions, Information extraction, Support vector machines",
author = "Stefania Rubrichi and Matteo Gabetta and Riccardo Bellazzi and Cristiana Larizza and Silvana Quaglini",
year = "2011",
language = "English",
volume = "761",
pages = "67--74",
booktitle = "CEUR Workshop Proceedings",

}

TY - GEN

T1 - Drug-Drug Interactions discovery based on CRFs, SVMs and rule-based methods

AU - Rubrichi, Stefania

AU - Gabetta, Matteo

AU - Bellazzi, Riccardo

AU - Larizza, Cristiana

AU - Quaglini, Silvana

PY - 2011

Y1 - 2011

N2 - Information about medications is critical in improving the patients' safety and quality of care. Most adverse drug events are predictable from the known pharmacology of the drugs and many represent known interactions and are, therefore, likely to be preventable. However, most of this information is locked in free-text and, as such, cannot be actively accessed and elaborated by computerized applications. In this work, we propose three different approaches to the problem of automatic recognition of drug-drug interactions that we have developed within the "First Challenge Task: Drug-Drug Interaction Extraction" competition. Our approaches learn to discriminate between semantically interesting and uninteresting content in a structured prediction framework as well as a rule-based one. The systems are trained using the DrugDDI corpus provided by the challenge organizers. An empirical analysis of the three approaches on this dataset shows that the inclusion of rule-based methods is indeed advantageous.

AB - Information about medications is critical in improving the patients' safety and quality of care. Most adverse drug events are predictable from the known pharmacology of the drugs and many represent known interactions and are, therefore, likely to be preventable. However, most of this information is locked in free-text and, as such, cannot be actively accessed and elaborated by computerized applications. In this work, we propose three different approaches to the problem of automatic recognition of drug-drug interactions that we have developed within the "First Challenge Task: Drug-Drug Interaction Extraction" competition. Our approaches learn to discriminate between semantically interesting and uninteresting content in a structured prediction framework as well as a rule-based one. The systems are trained using the DrugDDI corpus provided by the challenge organizers. An empirical analysis of the three approaches on this dataset shows that the inclusion of rule-based methods is indeed advantageous.

KW - Adverse drug events

KW - Conditional random fields

KW - Drug-Drug Interactions

KW - Information extraction

KW - Support vector machines

UR - http://www.scopus.com/inward/record.url?scp=84891111736&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84891111736&partnerID=8YFLogxK

M3 - Conference contribution

VL - 761

SP - 67

EP - 74

BT - CEUR Workshop Proceedings

ER -