In the rapidly evolving landscape of computer science and engineering, researchers have historically faced a binary choice: the speed and visibility of conferences or the depth and archival rigor of journals. Integrated publication frameworks—often implemented as "journal tracks"—aim to dissolve this dichotomy by creating a unified pipeline that offers the best of both worlds.
The Tradeoff Between Speed and Depth
Traditional scientific publishing operates on two different timescales. Conferences provide immediate community engagement and rapid dissemination, which is vital for fast-moving fields like machine learning and signal processing. However, the strict page limits and single-round "accept/reject" decisions of conferences can sometimes sacrifice technical depth.
Conversely, journals offer a multi-round, iterative peer-review process that ensures archival rigor. As noted by Karsten Borgwardt, Evimaria Terzi, and Po-Ling Loh in the Introduction to the special issue for the ECML PKDD 2019 journal track, the traditional model forces a tradeoff between "rapid dissemination" and "iterative journal feedback."
The Mechanism of the Journal Track
The "journal track" model functions as a hybrid dissemination pipeline. Instead of submitting a short paper to a conference, authors submit a full-length manuscript to a partner journal. If the paper passes the journal’s rigorous, multi-stage review process by a specific deadline, it is accepted for archival publication and the authors are invited to present their work at a major international conference.
This model has been successfully institutionalized across several major venues:
- ECML PKDD: Björn Bringmann and Élisa Fromont describe this in the Introduction to the special issue for the ECML PKDD 2018 journal track, highlighting how it allows "mature research to receive archival permanence" while maintaining the networking benefits of a live event.
- VLDB: The Proceedings of the VLDB Endowment, edited by Lei Chen and Fatma Ozcan, pioneered a year-round submission model that provides journal-style review with the guarantee of a presentation slot at the annual VLDB conference.
- ACL: The Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics integrated journal-published research to bridge the gap between long-form archival review and immediate community engagement.
Bidirectional Research Dissemination
Modern frameworks are increasingly "bidirectional." This means that not only do journal-track submissions get conference slots, but authors of regular journal papers are also invited to present at conferences to increase their work's reach.
Han Wei Shen and Kiyoshi Kiyokawa, writing in IEEE Transactions on Visualization and Computer Graphics, describe this as a "symbiotic relationship." By formalizing the presentation of journal-accepted work at venues like IEEE VR and ISMAR, the community ensures that the most thoroughly vetted research is also the most visible. This strategy fosters cross-pollination between archival research and active community discussion.
Why the Hybrid Model Works
The intuition behind this shift is that scientific value is maximized when high-quality vetting meets high-bandwidth communication. As Ira Assent and Carlotta Domeniconi argue in their editorial for ECML PKDD 2020, this approach ensures research meets "high standards of maturity and depth" without losing the "timeliness required in fast-moving technical fields."
Furthermore, as noted by Saeid Sanei regarding ICASSP 2019, these models provide a platform for real-time community feedback on work that has already undergone rigorous review. This creates a more robust scientific record where the community can discuss and build upon validated findings immediately.
Challenges in Evaluation
While these frameworks solve the dissemination problem, they introduce new challenges for peer review. In the Navigating Publication Outlets for Simulation Research panel, editors like Peter J. Haas and Christine Currie noted that as research becomes more interdisciplinary—integrating real-time data and machine learning—existing review frameworks must evolve to evaluate these hybrid methodologies effectively.
Ultimately, the goal of these integrated frameworks, as summarized by James Joshi in IEEE Transactions on Services Computing, is to strengthen the ties between archival venues and emerging research conferences, ensuring that the "freshest ideas" from the community are captured within a robust, permanent record.