Study on the Differences Between Journal Papers and Conference Papers in the Frontier of Basic Research: Taking the Terahertz Field as an Example

Hao Liu, Chen Yunwei, Biao Zhang2025 · DOI

TL;DR

By analyzing two decades of Terahertz (THz) research, this study reveals that journal papers serve as the primary backbone for theoretical breakthroughs and institutional collaboration, while conference papers act as the "fast lane" for technical and applied innovations. The findings highlight a significant geopolitical shift, with China overtaking the U.S. in THz publication volume since 2014, particularly in the journal space.

What Problem Does This Paper Address?

In the rapidly evolving "frontier" of basic research—fields characterized by high risk, high reward, and disruptive innovation—scientific output typically splits into journal articles and conference proceedings. However, it isn't always clear how these two mediums differ in their contribution to the field's evolution.

Do they cover the same topics? Do they foster the same level of international collaboration? The authors tackle these questions using the Terahertz field as a case study. Understanding these discrepancies is crucial for policymakers and researchers who need to identify where "disruptive" technologies are actually emerging and how knowledge flows through different academic channels.

Methods

The researchers employed a multi-pronged bibliometric and computational approach:

  • Social Network Analysis (SNA): They constructed institutional co-occurrence and citation networks. To compare these, they proposed a novel method to measure "network similarity" using PageRank (to evaluate node/institution importance) and Weight values (to evaluate the strength of connections/edges).
  • Text Mining (LDA): They used Latent Dirichlet Allocation (LDA) for topic clustering, analyzing titles and abstracts to identify eight distinct research themes within the THz domain.
  • Trend Analysis: They tracked publication volumes and geographical contributions over a 20-year period (2004–2023).

Datasets

The study utilized the Web of Science (WoS) Core Collection (specifically SCI-E and ESCI indices).

  • Scope: Papers published by IEEE between January 1, 2004, and December 31, 2023.
  • Volume: After deduplication, the dataset comprised 44,683 papers, consisting of 33,057 journal papers and 11,626 conference papers.

Key Insights and Ideas

  • The "Theory vs. Application" Divide: There is a clear functional split in the THz field. Theoretical topics like crystallography and optical materials are overwhelmingly published in journals (over 80%). Conversely, technical and applied topics like high-frequency communication systems and detectors have a much higher representation in conferences (up to 44%).
  • Journals as the Collaborative Hub: Perhaps the most surprising finding is that institutional collaboration is much more robust in journal publications. The similarity between the "journal-only" collaboration network and the "total" network was nearly 0.98, whereas the conference network sat around 0.75. This suggests that major institutional partnerships are primarily cemented through journal-grade research.
  • Geopolitical Momentum: The data quantifies a massive shift in THz leadership. In 2004, the U.S. led the field while China ranked 6th. By 2014, China surpassed the U.S. and has since maintained a dominant lead, particularly favoring journal publications (which account for 82% of China's THz output).
  • The "2020 Divergence": Around 2020, journal publications continued to climb while conference publications saw a decline. This likely reflects the impact of the COVID-19 pandemic on physical gatherings, but it also suggests a maturing field where researchers are prioritizing the "enduring influence" of journals over the "swift response" of conferences.

Key Results

  • Volume: Journal papers outnumber conference papers roughly 3 to 1 in this dataset.
  • Network Similarity: Journal citation networks showed a 0.999 similarity to the overall field's citation patterns, indicating they are the primary drivers of academic influence. Conference citation networks were significantly lower and experienced a sharp decline in similarity between 2014 and 2018.
  • Topic Hotspots: Terahertz optoelectronic radiation and optical materials are the largest current research clusters. However, high-frequency communication is identified as the fastest-growing "hotspot" for the near future.

Limitations and Future Work

The authors acknowledge several limitations:

  • Data Source Bias: The study only looked at IEEE publications within the WoS database. While IEEE is a giant in the THz field, it doesn't capture the entire global output.
  • Field Specificity: The patterns observed in Terahertz research (a mix of physics and engineering) might not apply to purely theoretical fields (like math) or purely applied fields (like software engineering).
  • Future Directions: The authors suggest that future work should expand the data scale and refine the methods for identifying "disruptive" technologies by specifically looking at the cross-pollination between conference-born ideas and their eventual journal-based maturation.

Summary updated

Abstract

Analyzing the discrepancies in the content of journal papers and conference papers in the frontier of basic research is beneficial for a comprehensive understanding of the characteristics and patterns of basic research development. Retrieve data from the Web of Science (WoS) database in the frontier of basic research. Employ a comprehensive approach using bibliometrics, social network analysis, and text mining methods to compare the differences in content between journal papers and conference papers.Explore aspects such as publication trends, paper contributions, and thematic evolution to analyze the disparities in the presentation of information between journals and conference proceedings.Terahertz crystallography, terahertz optical materials, and terahertz optoelectronic radiation tend to have theoretical research outcomes published more frequently in journal papers. On the other hand, terahertz high-frequency communication and application systems, terahertz communication technology, terahertz detectors, terahertz imaging, and measurement technology lean towards technical and applied research, with a preference for publication in conference papers.The research findings of this study uncover differences in literature characteristics and research topic across journal papers and conference paperss. This contributes to a more nuanced interpretation of the patterns in basic research development, ultimately enhancing the accuracy of identifying disruptive technologies and other related aspects in future investigations.
BibTeX
@misc{liu2025study,
  author = {Hao Liu and Chen Yunwei and Biao Zhang},
  title = {{Study on the Differences Between Journal Papers and Conference Papers in the Frontier of Basic Research: Taking the Terahertz Field as an Example}},
  year = {2025},
  doi = {10.51408/issi2025\_042},
}
Figure