Why Sociologists Should not Bother with Theory: The Effect of Topics on Citations

0510884 - FLÚ 2020 RIV IT eng C - Konferenční příspěvek (zahraniční konf.)
Hladík, Radim
Why Sociologists Should not Bother with Theory: The Effect of Topics on Citations.
17th International Conference on Scientometrics and Informetrics: ISSI 2019, Proceedings. Vol. 2. Rome: Sapienza University, 2019 - (Catalano, G.; Daraio, C.; Gregori, M.; Moed, H.; Ruocco, G.), s. 2341-2346. Efesto. ISBN 978-88-3381-118-5.
[International Conference on Scientometrics and Informetrics: ISSI 2019 /17./. Rome (IT), 02.09.2019-05.09.2019]
Institucionální podpora: RVO:67985955
Klíčová slova: Bibliometrics * Topic Modeling * Network-based * Odds ratios * Stochastic block models * Sociology * Czech
Obor OECD: Sociology
https://www.issi2019.org/

This paper utilizes odds ratios of topical words to model cumulative citation counts in a corpus of Czech sociological articles. It builds atop a novel method for topic modeling based on hierarchical stochastic block models and notes why network based topic models may be a more productive approach if the output of topic models is expected to serve as an input to further statistical models. The demonstrated technique shows that topics of articles have significant effects on citations and possibly interact with other important variables, such as the gender or number of authors.
Trvalý link: http://hdl.handle.net/11104/0301260