Citation networks have been widely used to study the evolution ofscience through the lenses of the underlying patterns of knowledge flowsamong academic papers, authors, research sub-fields, and scientificjournals. Here we focus on citation networks to cast light on thesalience of homophily, namely the principle that similarity breedsconnection, for knowledge transfer between papers. To this end, weassess the degree to which citations tend to occur between papers thatare concerned with seemingly related topics or research problems.Drawing on a large data set of articles published in the journals of theAmerican Physical Society between 1893 and 2009, we propose a novelmethod for measuring the similarity between articles through thestatistical validation of the overlap between their bibliographies.Results suggest that the probability of a citation made by one articleto another is indeed an increasing function of the similarity betweenthe two articles. Our study also enables us to uncover missing citationsbetween pairs of highly related articles, and may thus help identifybarriers to effective knowledge flows. By quantifying the proportion ofmissing citations, we conduct a comparative assessment of distinctjournals and research sub-fields in terms of their ability to facilitateor impede the dissemination of knowledge. Findings indicate thatElectromagnetism and Interdisciplinary Physics are the two sub-fields inphysics with the smallest percentage of missing citations. Moreover,knowledge transfer seems to be more effectively facilitated by journalsof wide visibility, such as Physical Review Letters, than bylower-impact ones. Our study has important implications for authors,editors and reviewers of scientific journals, as well as public preprintrepositories, as it provides a procedure for recommending relevant yetmissing references and properly integrating bibliographies of papers.

Homophily and missing links in citation networks

Vito Latora
2016-01-01

Abstract

Citation networks have been widely used to study the evolution ofscience through the lenses of the underlying patterns of knowledge flowsamong academic papers, authors, research sub-fields, and scientificjournals. Here we focus on citation networks to cast light on thesalience of homophily, namely the principle that similarity breedsconnection, for knowledge transfer between papers. To this end, weassess the degree to which citations tend to occur between papers thatare concerned with seemingly related topics or research problems.Drawing on a large data set of articles published in the journals of theAmerican Physical Society between 1893 and 2009, we propose a novelmethod for measuring the similarity between articles through thestatistical validation of the overlap between their bibliographies.Results suggest that the probability of a citation made by one articleto another is indeed an increasing function of the similarity betweenthe two articles. Our study also enables us to uncover missing citationsbetween pairs of highly related articles, and may thus help identifybarriers to effective knowledge flows. By quantifying the proportion ofmissing citations, we conduct a comparative assessment of distinctjournals and research sub-fields in terms of their ability to facilitateor impede the dissemination of knowledge. Findings indicate thatElectromagnetism and Interdisciplinary Physics are the two sub-fields inphysics with the smallest percentage of missing citations. Moreover,knowledge transfer seems to be more effectively facilitated by journalsof wide visibility, such as Physical Review Letters, than bylower-impact ones. Our study has important implications for authors,editors and reviewers of scientific journals, as well as public preprintrepositories, as it provides a procedure for recommending relevant yetmissing references and properly integrating bibliographies of papers.
2016
Citation networks, homophily, link prediction, bibliometric techniques.
File in questo prodotto:
File Dimensione Formato  
s13688-016-0068-2-homophily.pdf

accesso aperto

Tipologia: Versione Editoriale (PDF)
Dimensione 1.8 MB
Formato Adobe PDF
1.8 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11769/357676
Citazioni
  • ???jsp.display-item.citation.pmc??? 0
  • Scopus 37
  • ???jsp.display-item.citation.isi??? 35
social impact