Increasing NER recall with minimal precision loss

Jasper Kuperus, Maurice van Keulen, Cor J. Veenman

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

1 Citation (Scopus)
189 Downloads (Pure)

Abstract

Named Entity Recognition (NER) is broadly used as a first step toward the interpretation of text documents. However, for many applications, such as forensic investigation, recall is currently inadequate, leading to loss of potentially important information. Entity class ambiguity cannot be resolved reliably due to the lack of context information or the exploitation thereof. Consequently, entity classification introduces too many errors, leading to severe omissions in answers to forensic queries. We propose a technique based on multiple candidate labels effectively postponing decisions for entity classification to query time. Entity resolution exploits user feedback: a user is only asked for feedback on entities relevant to his/her query. Moreover, giving feedback can be stopped anytime when query results are considered good enough. We propose several interaction strategies that obtain increased recall with little loss in precision.
Original languageUndefined
Title of host publicationProceedings of the European Intelligence and Security Informatics Conference (EISIC 2013)
Place of PublicationUSA
PublisherIEEE Computer Society
Pages106-111
Number of pages6
ISBN (Print)978-0-7695-5062-6
DOIs
Publication statusPublished - Aug 2013

Publication series

Name
PublisherIEEE Computer Society

Keywords

  • EWI-23400
  • METIS-297674
  • IR-86348

Cite this

Kuperus, J., van Keulen, M., & Veenman, C. J. (2013). Increasing NER recall with minimal precision loss. In Proceedings of the European Intelligence and Security Informatics Conference (EISIC 2013) (pp. 106-111). USA: IEEE Computer Society. https://doi.org/10.1109/EISIC.2013.23
Kuperus, Jasper ; van Keulen, Maurice ; Veenman, Cor J. / Increasing NER recall with minimal precision loss. Proceedings of the European Intelligence and Security Informatics Conference (EISIC 2013). USA : IEEE Computer Society, 2013. pp. 106-111
@inproceedings{4e0eff33123b455a8aa04a53fe896c5b,
title = "Increasing NER recall with minimal precision loss",
abstract = "Named Entity Recognition (NER) is broadly used as a first step toward the interpretation of text documents. However, for many applications, such as forensic investigation, recall is currently inadequate, leading to loss of potentially important information. Entity class ambiguity cannot be resolved reliably due to the lack of context information or the exploitation thereof. Consequently, entity classification introduces too many errors, leading to severe omissions in answers to forensic queries. We propose a technique based on multiple candidate labels effectively postponing decisions for entity classification to query time. Entity resolution exploits user feedback: a user is only asked for feedback on entities relevant to his/her query. Moreover, giving feedback can be stopped anytime when query results are considered good enough. We propose several interaction strategies that obtain increased recall with little loss in precision.",
keywords = "EWI-23400, METIS-297674, IR-86348",
author = "Jasper Kuperus and {van Keulen}, Maurice and Veenman, {Cor J.}",
year = "2013",
month = "8",
doi = "10.1109/EISIC.2013.23",
language = "Undefined",
isbn = "978-0-7695-5062-6",
publisher = "IEEE Computer Society",
pages = "106--111",
booktitle = "Proceedings of the European Intelligence and Security Informatics Conference (EISIC 2013)",
address = "United States",

}

Kuperus, J, van Keulen, M & Veenman, CJ 2013, Increasing NER recall with minimal precision loss. in Proceedings of the European Intelligence and Security Informatics Conference (EISIC 2013). IEEE Computer Society, USA, pp. 106-111. https://doi.org/10.1109/EISIC.2013.23

Increasing NER recall with minimal precision loss. / Kuperus, Jasper; van Keulen, Maurice; Veenman, Cor J.

Proceedings of the European Intelligence and Security Informatics Conference (EISIC 2013). USA : IEEE Computer Society, 2013. p. 106-111.

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Increasing NER recall with minimal precision loss

AU - Kuperus, Jasper

AU - van Keulen, Maurice

AU - Veenman, Cor J.

PY - 2013/8

Y1 - 2013/8

N2 - Named Entity Recognition (NER) is broadly used as a first step toward the interpretation of text documents. However, for many applications, such as forensic investigation, recall is currently inadequate, leading to loss of potentially important information. Entity class ambiguity cannot be resolved reliably due to the lack of context information or the exploitation thereof. Consequently, entity classification introduces too many errors, leading to severe omissions in answers to forensic queries. We propose a technique based on multiple candidate labels effectively postponing decisions for entity classification to query time. Entity resolution exploits user feedback: a user is only asked for feedback on entities relevant to his/her query. Moreover, giving feedback can be stopped anytime when query results are considered good enough. We propose several interaction strategies that obtain increased recall with little loss in precision.

AB - Named Entity Recognition (NER) is broadly used as a first step toward the interpretation of text documents. However, for many applications, such as forensic investigation, recall is currently inadequate, leading to loss of potentially important information. Entity class ambiguity cannot be resolved reliably due to the lack of context information or the exploitation thereof. Consequently, entity classification introduces too many errors, leading to severe omissions in answers to forensic queries. We propose a technique based on multiple candidate labels effectively postponing decisions for entity classification to query time. Entity resolution exploits user feedback: a user is only asked for feedback on entities relevant to his/her query. Moreover, giving feedback can be stopped anytime when query results are considered good enough. We propose several interaction strategies that obtain increased recall with little loss in precision.

KW - EWI-23400

KW - METIS-297674

KW - IR-86348

U2 - 10.1109/EISIC.2013.23

DO - 10.1109/EISIC.2013.23

M3 - Conference contribution

SN - 978-0-7695-5062-6

SP - 106

EP - 111

BT - Proceedings of the European Intelligence and Security Informatics Conference (EISIC 2013)

PB - IEEE Computer Society

CY - USA

ER -

Kuperus J, van Keulen M, Veenman CJ. Increasing NER recall with minimal precision loss. In Proceedings of the European Intelligence and Security Informatics Conference (EISIC 2013). USA: IEEE Computer Society. 2013. p. 106-111 https://doi.org/10.1109/EISIC.2013.23