Recommending collaborators using social features and MeSH terms

Danielle H. Lee, Peter Brusilovsky, Titus Schleyer

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

Unlike expertise location systems which users query actively when looking for an expert, expert recommender systems suggest individuals without the context of a specific problem. An interesting research question is whether expert recommender systems should consider a users' social context when recommending potential research collaborators. One may argue that it might be easier for scientists to collaborate with colleagues in their social network, because initiating collaboration with socially unconnected researchers is burdensome and fraught with risk, despite potentially relevant expertise. However, many scientists also initiate collaborations outside of their social network when they seek to work with individuals possessing relevant expertise or acknowledged experts. In this paper, we studied how well content-based, social and hybrid recommendation algorithms predicted co-author relationships among a random sample of 17,525 biomedical scientists. To generate recommendations, we used authors' research expertise inferred from publication metadata and their professional social networks derived from their coauthorship history. We used 80% of our data set (articles published before 2007) as our training set, and the remaining data as our test set (articles published in 2007 or later). Our results show that a hybrid algorithm combining expertise and social network information outperformed all other algorithms with regards to Top 10 and Top 20 recommendations. For the Top 2 and Top 5 recommendations, social network-based information alone generated the most useful recommendations. Our study provides evidence that integrating social network information in expert recommendations may outperform a purely expertise-based approach.

Original languageEnglish (US)
JournalProceedings of the ASIST Annual Meeting
Volume48
DOIs
StatePublished - 2011
Externally publishedYes

Fingerprint

social network
expertise
Recommender systems
knowledge-based system
expert
Metadata
random sample
history
evidence

Keywords

  • Biomedical research
  • Expert recommendation
  • Medical subject headings (MeSH)
  • Social networks

ASJC Scopus subject areas

  • Information Systems
  • Library and Information Sciences

Cite this

Recommending collaborators using social features and MeSH terms. / Lee, Danielle H.; Brusilovsky, Peter; Schleyer, Titus.

In: Proceedings of the ASIST Annual Meeting, Vol. 48, 2011.

Research output: Contribution to journalArticle

@article{d7653cf6c0e948479725a3d817cf1348,
title = "Recommending collaborators using social features and MeSH terms",
abstract = "Unlike expertise location systems which users query actively when looking for an expert, expert recommender systems suggest individuals without the context of a specific problem. An interesting research question is whether expert recommender systems should consider a users' social context when recommending potential research collaborators. One may argue that it might be easier for scientists to collaborate with colleagues in their social network, because initiating collaboration with socially unconnected researchers is burdensome and fraught with risk, despite potentially relevant expertise. However, many scientists also initiate collaborations outside of their social network when they seek to work with individuals possessing relevant expertise or acknowledged experts. In this paper, we studied how well content-based, social and hybrid recommendation algorithms predicted co-author relationships among a random sample of 17,525 biomedical scientists. To generate recommendations, we used authors' research expertise inferred from publication metadata and their professional social networks derived from their coauthorship history. We used 80{\%} of our data set (articles published before 2007) as our training set, and the remaining data as our test set (articles published in 2007 or later). Our results show that a hybrid algorithm combining expertise and social network information outperformed all other algorithms with regards to Top 10 and Top 20 recommendations. For the Top 2 and Top 5 recommendations, social network-based information alone generated the most useful recommendations. Our study provides evidence that integrating social network information in expert recommendations may outperform a purely expertise-based approach.",
keywords = "Biomedical research, Expert recommendation, Medical subject headings (MeSH), Social networks",
author = "Lee, {Danielle H.} and Peter Brusilovsky and Titus Schleyer",
year = "2011",
doi = "10.1002/meet.2011.14504801025",
language = "English (US)",
volume = "48",
journal = "Proceedings of the ASIST Annual Meeting",
issn = "1550-8390",
publisher = "Learned Information",

}

TY - JOUR

T1 - Recommending collaborators using social features and MeSH terms

AU - Lee, Danielle H.

AU - Brusilovsky, Peter

AU - Schleyer, Titus

PY - 2011

Y1 - 2011

N2 - Unlike expertise location systems which users query actively when looking for an expert, expert recommender systems suggest individuals without the context of a specific problem. An interesting research question is whether expert recommender systems should consider a users' social context when recommending potential research collaborators. One may argue that it might be easier for scientists to collaborate with colleagues in their social network, because initiating collaboration with socially unconnected researchers is burdensome and fraught with risk, despite potentially relevant expertise. However, many scientists also initiate collaborations outside of their social network when they seek to work with individuals possessing relevant expertise or acknowledged experts. In this paper, we studied how well content-based, social and hybrid recommendation algorithms predicted co-author relationships among a random sample of 17,525 biomedical scientists. To generate recommendations, we used authors' research expertise inferred from publication metadata and their professional social networks derived from their coauthorship history. We used 80% of our data set (articles published before 2007) as our training set, and the remaining data as our test set (articles published in 2007 or later). Our results show that a hybrid algorithm combining expertise and social network information outperformed all other algorithms with regards to Top 10 and Top 20 recommendations. For the Top 2 and Top 5 recommendations, social network-based information alone generated the most useful recommendations. Our study provides evidence that integrating social network information in expert recommendations may outperform a purely expertise-based approach.

AB - Unlike expertise location systems which users query actively when looking for an expert, expert recommender systems suggest individuals without the context of a specific problem. An interesting research question is whether expert recommender systems should consider a users' social context when recommending potential research collaborators. One may argue that it might be easier for scientists to collaborate with colleagues in their social network, because initiating collaboration with socially unconnected researchers is burdensome and fraught with risk, despite potentially relevant expertise. However, many scientists also initiate collaborations outside of their social network when they seek to work with individuals possessing relevant expertise or acknowledged experts. In this paper, we studied how well content-based, social and hybrid recommendation algorithms predicted co-author relationships among a random sample of 17,525 biomedical scientists. To generate recommendations, we used authors' research expertise inferred from publication metadata and their professional social networks derived from their coauthorship history. We used 80% of our data set (articles published before 2007) as our training set, and the remaining data as our test set (articles published in 2007 or later). Our results show that a hybrid algorithm combining expertise and social network information outperformed all other algorithms with regards to Top 10 and Top 20 recommendations. For the Top 2 and Top 5 recommendations, social network-based information alone generated the most useful recommendations. Our study provides evidence that integrating social network information in expert recommendations may outperform a purely expertise-based approach.

KW - Biomedical research

KW - Expert recommendation

KW - Medical subject headings (MeSH)

KW - Social networks

UR - http://www.scopus.com/inward/record.url?scp=84861434329&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84861434329&partnerID=8YFLogxK

U2 - 10.1002/meet.2011.14504801025

DO - 10.1002/meet.2011.14504801025

M3 - Article

AN - SCOPUS:84861434329

VL - 48

JO - Proceedings of the ASIST Annual Meeting

JF - Proceedings of the ASIST Annual Meeting

SN - 1550-8390

ER -