Exploring author name disambiguation on PubMed-scale

Min Song, Erin Hea Jin Kim, Ha Jin Kim

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Author name disambiguation (AND) creates a daunting challenge in that disambiguation techniques often draw false conclusions when applied to incomplete or incorrect publication data. It becomes a more critical issue in the biomedical domain where PubMed articles are written by a wide range of researchers internationally. To tackle this issue, we create a carefully hand-crafted training set drawn from the entire PubMed collection by going through multiple iterations. We assess the quality of our training set by comparing it with SCOPUS-based training set. In addition, for the performance enhancement of the AND techniques, we propose a new set of publication features extracted by text mining techniques. The results of the experiments show that all four supervised learning techniques (Random Forest, C4.5, KNN, and SVM) with the new publication features (called NER model) achieve improved performance over the baseline and hybrid edit distance model.

Original languageEnglish
Pages (from-to)924-941
Number of pages18
JournalJournal of Informetrics
Volume9
Issue number4
DOIs
Publication statusPublished - 2015 Oct 1

Fingerprint

Supervised learning
performance
Experiments
experiment
learning

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Library and Information Sciences

Cite this

Song, Min ; Kim, Erin Hea Jin ; Kim, Ha Jin. / Exploring author name disambiguation on PubMed-scale. In: Journal of Informetrics. 2015 ; Vol. 9, No. 4. pp. 924-941.
@article{d7ff279ab8e94f598ba6a62d89579cec,
title = "Exploring author name disambiguation on PubMed-scale",
abstract = "Author name disambiguation (AND) creates a daunting challenge in that disambiguation techniques often draw false conclusions when applied to incomplete or incorrect publication data. It becomes a more critical issue in the biomedical domain where PubMed articles are written by a wide range of researchers internationally. To tackle this issue, we create a carefully hand-crafted training set drawn from the entire PubMed collection by going through multiple iterations. We assess the quality of our training set by comparing it with SCOPUS-based training set. In addition, for the performance enhancement of the AND techniques, we propose a new set of publication features extracted by text mining techniques. The results of the experiments show that all four supervised learning techniques (Random Forest, C4.5, KNN, and SVM) with the new publication features (called NER model) achieve improved performance over the baseline and hybrid edit distance model.",
author = "Min Song and Kim, {Erin Hea Jin} and Kim, {Ha Jin}",
year = "2015",
month = "10",
day = "1",
doi = "10.1016/j.joi.2015.08.004",
language = "English",
volume = "9",
pages = "924--941",
journal = "Journal of Informetrics",
issn = "1751-1577",
publisher = "Elsevier BV",
number = "4",

}

Exploring author name disambiguation on PubMed-scale. / Song, Min; Kim, Erin Hea Jin; Kim, Ha Jin.

In: Journal of Informetrics, Vol. 9, No. 4, 01.10.2015, p. 924-941.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Exploring author name disambiguation on PubMed-scale

AU - Song, Min

AU - Kim, Erin Hea Jin

AU - Kim, Ha Jin

PY - 2015/10/1

Y1 - 2015/10/1

N2 - Author name disambiguation (AND) creates a daunting challenge in that disambiguation techniques often draw false conclusions when applied to incomplete or incorrect publication data. It becomes a more critical issue in the biomedical domain where PubMed articles are written by a wide range of researchers internationally. To tackle this issue, we create a carefully hand-crafted training set drawn from the entire PubMed collection by going through multiple iterations. We assess the quality of our training set by comparing it with SCOPUS-based training set. In addition, for the performance enhancement of the AND techniques, we propose a new set of publication features extracted by text mining techniques. The results of the experiments show that all four supervised learning techniques (Random Forest, C4.5, KNN, and SVM) with the new publication features (called NER model) achieve improved performance over the baseline and hybrid edit distance model.

AB - Author name disambiguation (AND) creates a daunting challenge in that disambiguation techniques often draw false conclusions when applied to incomplete or incorrect publication data. It becomes a more critical issue in the biomedical domain where PubMed articles are written by a wide range of researchers internationally. To tackle this issue, we create a carefully hand-crafted training set drawn from the entire PubMed collection by going through multiple iterations. We assess the quality of our training set by comparing it with SCOPUS-based training set. In addition, for the performance enhancement of the AND techniques, we propose a new set of publication features extracted by text mining techniques. The results of the experiments show that all four supervised learning techniques (Random Forest, C4.5, KNN, and SVM) with the new publication features (called NER model) achieve improved performance over the baseline and hybrid edit distance model.

UR - http://www.scopus.com/inward/record.url?scp=84944182092&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84944182092&partnerID=8YFLogxK

U2 - 10.1016/j.joi.2015.08.004

DO - 10.1016/j.joi.2015.08.004

M3 - Article

VL - 9

SP - 924

EP - 941

JO - Journal of Informetrics

JF - Journal of Informetrics

SN - 1751-1577

IS - 4

ER -