Partners in Criminology: Machine Learning and Network Science Reveal Missed Opportunities and Inequalities in the Study of Crime

ObjectivesResearch in criminology and criminal justice is a rapidly growing interdisciplinary, international network of intersecting research topics. Quantifying the substantive content of criminology allows us to empirically disentangle this network and analyze how the interplay of research topic a...

Ausführliche Beschreibung

Gespeichert in:  
Bibliographische Detailangaben
1. VerfasserIn: Smith, Thomas Bryan (VerfasserIn)
Beteiligte: Mao, Ruijie ; Korotchenko, Stan ; Krohn, Marvin D. 1947-
Medienart: Elektronisch Aufsatz
Sprache:Englisch
Veröffentlicht: 2024
In: Journal of quantitative criminology
Jahr: 2024, Band: 40, Heft: 2, Seiten: 421-443
Online-Zugang: Volltext (lizenzpflichtig)
Journals Online & Print:
Lade...
Verfügbarkeit prüfen: HBZ Gateway
Schlagwörter:
Asian Criminology
Biosocial Criminology
Coauthorship
Criminology Theory
Critical Criminology
machine learning
Quantitative Criminology
Research Methods in Criminology
Science of science
Social Network Analysis
Topic modelling
Beschreibung
Zusammenfassung:ObjectivesResearch in criminology and criminal justice is a rapidly growing interdisciplinary, international network of intersecting research topics. Quantifying the substantive content of criminology allows us to empirically disentangle this network and analyze how the interplay of research topic and international context influences knowledge production in the study of crime.MethodsIn this paper we apply the topic embedding model, top2vec, mapping two decades of research in criminology and criminal justice (2001-2020). Using data generated by top2vec we enrich coauthorship network data—introducing themes and subdisciplines, and topical similarity measures—and analyze the relationship between topical and subdisciplinary overlap, geospatial distance, and publication coauthorship in a sample of mid-career researchers (N = 4068).ResultsWe find that these researchers disproportionately favor collaboration within their immediate network of collaborators and subdiscipline of criminology, but tend to establish new, synergistic collaborations with topically adjacent researchers. New collaborations appear to be independent of geographic distance, but US criminologists are less likely to collaborate internationally than Eurasian and Oceanic criminologists.ConclusionsFacilitating communication between researchers and organizations from adjacent subdisciplines could benefit researchers and produce new, innovative research. Encouraging comparative research would help international scholars, many of whom may rely on US collaborators, benefit more from US scholarship. Further applications for top2vec in the scientometric study of criminology, criminal justice, and legal studies are discussed.
ISSN:1573-7799
DOI:10.1007/s10940-023-09574-z