Automatic classification of PubMed abstracts with Latent semantic indexing

Working notes

Joel Robert Adams, Steven Bedrick

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

Abstract

The 2014 BioASQ challenge 2a tasks participants with assigning semantic tags to biomedical journal abstracts. We present a system that uses Latent Semantic Analysis to identify semantically similar documents in MEDLINE to an unlabeled abstract, and then uses a novel ranking scheme to select a list of MeSH headers from candidates drawn from the most similar documents. Our approach achieved good precision, but suffered in terms of recall. We describe several possible strategies to improve our system's performance.

Original languageEnglish (US)
Title of host publicationCLEF 2014 - Working Notes for CLEF 2014 Conference
PublisherCEUR-WS
Pages1275-1282
Number of pages8
Volume1180
StatePublished - 2014
Event2014 Cross Language Evaluation Forum Conference, CLEF 2014 - Sheffield, United Kingdom
Duration: Sep 15 2014Sep 18 2014

Other

Other2014 Cross Language Evaluation Forum Conference, CLEF 2014
CountryUnited Kingdom
CitySheffield
Period9/15/149/18/14

Fingerprint

Semantics

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Adams, J. R., & Bedrick, S. (2014). Automatic classification of PubMed abstracts with Latent semantic indexing: Working notes. In CLEF 2014 - Working Notes for CLEF 2014 Conference (Vol. 1180, pp. 1275-1282). CEUR-WS.

Automatic classification of PubMed abstracts with Latent semantic indexing : Working notes. / Adams, Joel Robert; Bedrick, Steven.

CLEF 2014 - Working Notes for CLEF 2014 Conference. Vol. 1180 CEUR-WS, 2014. p. 1275-1282.

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

Adams, JR & Bedrick, S 2014, Automatic classification of PubMed abstracts with Latent semantic indexing: Working notes. in CLEF 2014 - Working Notes for CLEF 2014 Conference. vol. 1180, CEUR-WS, pp. 1275-1282, 2014 Cross Language Evaluation Forum Conference, CLEF 2014, Sheffield, United Kingdom, 9/15/14.
Adams JR, Bedrick S. Automatic classification of PubMed abstracts with Latent semantic indexing: Working notes. In CLEF 2014 - Working Notes for CLEF 2014 Conference. Vol. 1180. CEUR-WS. 2014. p. 1275-1282
Adams, Joel Robert ; Bedrick, Steven. / Automatic classification of PubMed abstracts with Latent semantic indexing : Working notes. CLEF 2014 - Working Notes for CLEF 2014 Conference. Vol. 1180 CEUR-WS, 2014. pp. 1275-1282
@inproceedings{1bf2d9f2951349e8ac2e6acc41a526aa,
title = "Automatic classification of PubMed abstracts with Latent semantic indexing: Working notes",
abstract = "The 2014 BioASQ challenge 2a tasks participants with assigning semantic tags to biomedical journal abstracts. We present a system that uses Latent Semantic Analysis to identify semantically similar documents in MEDLINE to an unlabeled abstract, and then uses a novel ranking scheme to select a list of MeSH headers from candidates drawn from the most similar documents. Our approach achieved good precision, but suffered in terms of recall. We describe several possible strategies to improve our system's performance.",
author = "Adams, {Joel Robert} and Steven Bedrick",
year = "2014",
language = "English (US)",
volume = "1180",
pages = "1275--1282",
booktitle = "CLEF 2014 - Working Notes for CLEF 2014 Conference",
publisher = "CEUR-WS",

}

TY - GEN

T1 - Automatic classification of PubMed abstracts with Latent semantic indexing

T2 - Working notes

AU - Adams, Joel Robert

AU - Bedrick, Steven

PY - 2014

Y1 - 2014

N2 - The 2014 BioASQ challenge 2a tasks participants with assigning semantic tags to biomedical journal abstracts. We present a system that uses Latent Semantic Analysis to identify semantically similar documents in MEDLINE to an unlabeled abstract, and then uses a novel ranking scheme to select a list of MeSH headers from candidates drawn from the most similar documents. Our approach achieved good precision, but suffered in terms of recall. We describe several possible strategies to improve our system's performance.

AB - The 2014 BioASQ challenge 2a tasks participants with assigning semantic tags to biomedical journal abstracts. We present a system that uses Latent Semantic Analysis to identify semantically similar documents in MEDLINE to an unlabeled abstract, and then uses a novel ranking scheme to select a list of MeSH headers from candidates drawn from the most similar documents. Our approach achieved good precision, but suffered in terms of recall. We describe several possible strategies to improve our system's performance.

UR - http://www.scopus.com/inward/record.url?scp=84961355834&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84961355834&partnerID=8YFLogxK

M3 - Conference contribution

VL - 1180

SP - 1275

EP - 1282

BT - CLEF 2014 - Working Notes for CLEF 2014 Conference

PB - CEUR-WS

ER -