Time-sensitive research through preprints: employing topic modeling based on correlation and significance

Tzu-Yu Lin and Pei-Chun Lee ()
Additional contact information
Tzu-Yu Lin: National Chengchi University, Graduate Institute of Library, Information and Archival Studies
Pei-Chun Lee: National Chengchi University, Graduate Institute of Library, Information and Archival Studies

Scientometrics, 2025, vol. 130, issue 10, No 11, 5547-5569

Abstract: Abstract In rapidly evolving scientific domains such as edge computing, timely identification of emerging research directions is increasingly critical. Preprints, as fast and openly accessible scholarly outputs, have long served as an important medium for the early dissemination of scholarly knowledge, particularly in high-velocity fields. However, their lack of peer review and heterogeneous quality present challenges for systematic analysis. This study examines edge computing preprints published between 2021 and 2023, retrieved from the Web of Science Preprint Citation Index. A hybrid topic modeling approach combining Latent Dirichlet Allocation (LDA) and Word2Vec-based semantic similarity was applied, along with an institutional credibility–based filtering mechanism to ensure corpus quality. The findings reveal three prominent thematic clusters: (1) resource management and task scheduling, focusing on optimizing computational infrastructure; (2) edge intelligence and federated learning, highlighting distributed AI applications; and (3) real-time analytics and low-latency processing, reflecting network efficiency concerns. These results align with prior research and suggest that preprints can complement traditional literature in capturing early research signals. The proposed framework demonstrates adaptability to dynamic, unstructured corpora and offers a scalable method for real-time knowledge monitoring in emerging technological fields.

Keywords: Preprints; Topic modeling; Latent Dirichlet allocation; Word embedding; Edge computing (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11192-025-05437-5 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:scient:v:130:y:2025:i:10:d:10.1007_s11192-025-05437-5

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11192

DOI: 10.1007/s11192-025-05437-5

Access Statistics for this article

Scientometrics is currently edited by Wolfgang Glänzel

More articles in Scientometrics from Springer, Akadémiai Kiadó
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().