Named entity recognition in chemical patents using ensemble of contextual language models

Copara, Jenny (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; Swiss Institute of Bioinformatics, Geneva, Switzerland ; University of Geneva, Switzerland) ; Naderi, Nona (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; Swiss Institute of Bioinformatics, Geneva, Switzerland) ; Knafou, Julien (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; Swiss Institute of Bioinformatics, Geneva, Switzerland ; University of Geneva, Switzerland) ; Ruch, Patrick (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; Swiss Institute of Bioinformatics, Geneva, Switzerland) ; Teodoro, Douglas (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; Swiss Institute of Bioinformatics, Geneva, Switzerland)

Chemical patent documents describe a broad range of applications holding key reaction and compound information, such as chemical structure, reaction formulas, and molecular properties. These informational entities should be first identifed in text passages to be utilized in downstream tasks. Text mining provides means to extract relevant information from chemical patents through information extraction techniques. As part of the Information Extraction task of the Cheminformatics Elsevier Melbourne University challenge, in this work we study the effectiveness of contextualized language models to extract reaction information in chemical patents. We assess transformer architectures trained on a generic and specialised corpora to propose a new ensemble model. Our best model, based on a majority ensemble approach, achieves an exact F1-score of 92:30% and a relaxed F1-score of 96:24%. The results show that ensemble of contextualized language models can provide an effective method to extract information from chemical patents.


Note: Due to the COVID-19 outbreak, the CLEF 2020 conference venue in Thessaloniki was cancelled. The proceedings of the online conference are however published according to the original schedule.


Keywords:
Conference Type:
full paper
Faculty:
Economie et Services
School:
HEG - Genève
Institute:
CRAG - Centre de Recherche Appliquée en Gestion
Subject(s):
Informatique
Publisher:
Thessaloniki, Greece, 22-25 September 2020
Date:
2020-09
Thessaloniki, Greece
22-25 September 2020
Pagination:
15 p.
Published in:
Proceedings of the CLEF 2020 conference
External resources:
Appears in Collection:



 Record created 2020-10-05, last modified 2020-10-28

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)