Goals

Data visualization requires a thoughtful design process that heavily relies on both domain knowledge and familiarity with visualization techniques. Given the vast design space and inherent complexity, even experts often invest substantial effort to create effective visualizations for exploration or communication.

With the rapid advancement of artificial intelligence (AI)—particularly the rise of powerful foundation models such as large language models (LLMs), vision-language models (VLMs), and multimodal AI systems—the field of visualization is undergoing a significant transformation. These cutting-edge models present new opportunities to automate and augment the visualization process. For instance, LLMs can translate natural language queries into visual specifications, assist with data wrangling, and recommend appropriate visualization types. VLMs and multimodal models enable deeper data understanding and interaction by integrating textual, visual, and tabular information, further advancing the creation of intuitive and intelligent visualizations.

Concurrently, visualization plays an increasingly vital role in the development and deployment of advanced AI models. As these models grow in complexity and scale, the need for effective visual interfaces and techniques becomes more pressing—to interpret model behavior, debug outputs, and foster transparency, accountability, and human-AI collaboration.

This workshop, held in conjunction with IEEE PacificVis 2027, aims to explore this dynamic and rapidly evolving area by fostering communication between the visualization and AI communities. Attendees will engage with the latest research at the intersection of AI-enhanced visualization (AI4VIS) and visualization-enhanced AI (VIS4AI), with a particular focus on how cutting-edge models—such as LLMs, VLMs, and beyond—are reshaping the landscape.

Call for Participation

Submission

We welcome submissions in the form of full papers. All accepted papers will be published in a special issue of the Information Visualization journal.

Authors should follow the instructions under "Preparing your manuscript for submission" as outlined in the journal's submission guidelines. A LaTeX template is available under the "Article format" section at this link. While there is no strict page limit, authors are encouraged to ensure that the length of their paper appropriately reflects the scope and significance of their contribution. For reference, a typical full paper using the Information Visualization LaTeX template (double-column) is approximately 15–20 pages. If you have any questions about this guideline, please feel free to contact the chairs.

Submissions must be made through PCS (Track Name: PacificVis 2027 Visualization Meets AI Workshop). Only double-blind (anonymized) submissions will be accepted. Please replace author names with the paper ID number to ensure anonymity.

Topics of Interest

We invite high-quality research and application papers that integrate visualization and AI/machine learning. Submissions may address AI for Visualization (AI4VIS), Visualization for AI (VIS4AI), or both.

The following are representative papers from previous editions of the workshop. For additional examples, please explore the latest special issue of Information Visualization on Visualization Meets AI.

AI4VIS
S. Jung, J. Rhee, S. Doh, H. Jeon, G. J. Quadri, and J. Seo. Seeing Graphs Like Humans: Benchmarking Computational Measures and MLLMs for Similarity Assessment. Information Visualization, 25(3): 333-357, 2026.
G. Zhao, Z. Wang, Y. Dong, G. Li, and G. Shan Toward Reliable Scientific Visualization Pipeline Construction with Structure-Aware Retrieval-Augmented LLMs. Information Visualization, 25(3), 373-390, 2026.
P.-P. Vázquez. Are LLMs ready for Visualization? In 2024 IEEE 17th Pacific Visualization Conference (PacificVis), pp. 343-352, 2024.
J. Han and C. Wang. VCNet: A Generative Model for Volume Completion. Visual Informatics, 6(2): 62-73, 2022.
L. Giovannangeli, R. Bourqui, R. Giot, and D. Auber. Toward Automatic Comparison of Visualization Techniques: Application to Graph Visualization. Visual Informatics, 4(2): 86-98, 2020.
VIS4AI
D. Collaris, Y. Liang, M. C. Willemsen, A. Chatzimparmpas, and J. J. van Wijk. Evaluating the Utility of Feature Importance Visualizations in SHAP. Information Visualization, 25(3): 391-409, 2026.
X. Liang, Y. Chen, and C. Lv. DAttnVis: Attention-Guided Visual Diagnostics for Stable Diffusion Inference in Image Generation. Information Visualization, 25(3), 268-287, 2026.
Z. Liang, G. Li, R. Gu, Y. Wang, and G. Shan. SampleViz: Concept based Sampling for Policy Refinement in Deep Reinforcement Learning. In 2024 IEEE 17th Pacific Visualization Conference (PacificVis), pp. 359-368, 2024.
M. Gleicher, X. Yu, and Y. Chen. Trinary Tools for Continuously Valued Binary Classifiers. Visual Informatics, 6(2): 74-86, 2022.
X. Ji, Y. Tu, W. He, J. Wang, H.-W. Shen, and P.-Y. Yen. USEVis: Visual Analytics of Attention-Based Neural Embedding in Information Retrieval. Visual Informatics, 5(2): 1-12, 2021.

Important Dates

December 18, 2026: Paper due
February 1, 2027: 1st cycle notification from workshop chairs
Conditionally accepted papers need to go through minor revisions and to be reviewed in the second review cycle. This year, we will also recommend promising papers with strong potential, but not yet ready for acceptance, for the fast-track review process at Information Visualization. In this process, we will try our best to maintain reviewer continuity whenever possible.
February 15, 2027: Revision due
February 22, 2027: 2nd cycle notification from workshop chairs
Workshop chairs will recommend acceptance to the Editor-in-Chief (EIC) of Information Visualization if the revisions are deemed satisfactory. However, papers with insufficient revisions may still be rejected during this cycle or moved to the fast-track review process.
March 1, 2027: Editable source files due
At this stage, authors should avoid making major changes, such as modifying the paper title or author information. Such changes may delay the production process and could result in the paper being declined for publication.
March 8, 2027: Final notification from the EIC of Information Visualization
April 19, 2027: Workshop
Accepted papers may still be in production at the time of the workshop and may appear online afterward.

All deadlines are due at 11:59pm (23:59) Anywhere on Earth (AoE).

Committees

Workshop Chair

Takanori Fujiwara

Takanori Fujiwara

University of Arizona

Junpeng Wang

Junpeng Wang

Visa Research

Program Committee

Coming soon

Past Events

Contact

pvis_ai4vis@pvis.org