Accepted Submissions
GUGOG @ IEEE VIS 2026
Seven short position papers were accepted to GUGOG. Camera-ready versions are hosted here. Submissions are non-archival, and copyright remains with the authors.
Presented at the Workshop
Five of the accepted papers will be presented as talks on November 9, 2026 (see the schedule).
Mind Your νs and Qs: A Categorical Framework for Specifying Visualization Components
Abstract
Graphical grammars express how to put visual elements together; we propose a methodology for defining those visual elements in a manner that guarantees that they are structure preserving. Fiber bundles provide a uniform interface for describing a variety of data (tables, images, networks, etc.) in a manner that independently encodes the mathematical structure of the continuity and fields of the dataset. Modeling the data containers that store the datasets as sheaves provides a framework for specifying visualization methods that are designed to work regardless of how the dataset is stored, whether on disk, distributed, or on demand. Specifying the visualization library components as natural transforms of sheaves means that the constraints that the component must satisfy to be structure preserving can be specified as the set of morphisms on the data and graphic sheaves, including the structure on the continuity and fields of the data. Formally expressing how visual elements are constructed using category theory means we can translate those constraints into pseudocode, which can be used to implement the words in the visual sentence in a way that ensures that the sentences generate images that are faithful translations of the data.
On the Inconsistency in Grammar Evaluations
Abstract
Current evaluation methods for visualization grammars are fragmented and inadequate. This work explores the community's lack of a standardized approach to evaluating grammars and comparing them. We propose broadening the evaluation criteria to include metrics like ease of use, conciseness, expressivity, and more rigorous evaluation techniques.
Balancing constriction and expansion for domain-specific grammars
Abstract
Graphical grammars distinguish from heuristic typologies in constituting closed formal systems whose product diversity arises from recombination rather than from de novo redesign. In practice, the scope of an implementation is constrained both by what options have been implemented for each component (e.g. the polar coordinatization, the grouped count statistic, and the rectangle geometric element) and by what combinations are enabled (these three yielding a pie graphic). In the interest of flexibility, implementations are generally permissive of syntactically valid combinations. However, within some domains, universal or conditional rules may be in place to ensure that graphics faithfully communicate important properties of the input data, such that permitting violations would compromise their integrity. I will present geometric data analysis as a motivating case study for this situation. Visualizations of point clouds, and especially the use of biplots, relies universally on a unit aspect ratio and conditionally on coordinate standardization for plot elements to approximate geometric relationships among data elements. Rather than a limitation of graphical grammar, these should be viewed as opportunities to explore the phrase spaces of different graphical domains. Indeed, the geometric constraints above enable the meaningful use of two families of statistical transformation (projections of constructions in higher dimensions and constructions in relation to referent points) that would otherwise produce uninterpretable graphics. This leads me to argue for a change in language from Wilkinson's assertion that, while a grammatical system can produce hideous graphics, it cannot produce meaningless ones. I suggest that meaningfulness and meaninglessness are not mathematical properties secured by formal rules but semantic properties vested with cultural significance, and that graphical grammars must adapt to the culture of a domain in order to serve as trustworthy tools for communicating information within it.
The Grammar of Graphics Needs Verbs: Towards a GUGOG
Abstract
The Grammar of Graphics (GoG) made significant strides toward unifying previously incompatible chart abstractions into a shared conceptual model. Today, we face a related problem. Inspired by the GoG, we have a proliferation of domain-specific grammars whose compositional ideas are difficult to compare and combine. I argue that one source of this divergence is that while the GoG gives us a powerful set of nouns for statistical graphics (marks), it provides no equally general set of verbs for composing marks together. Through this lens, we can understand recent grammars as specialized extensions of the GoG with verbs for arranging, compositing, repeating, or otherwise relating marks to each other. But because each grammar introduces these operations in its own language, closely related ideas are still difficult to compare, combine, and carry into practice. In this position paper, I show how the GoFish grammar's relations and operators can represent three recent grammars in a shared system. GoFish allows the same verbs to be used across all three systems, making it easier to compare and contrast grammars and identify their core contributions.
Semantic Validity vs. Graphical Arrangement in Visualization Grammars
Abstract
Systems for modular, declarative and compositional specification for data visualizations, otherwise known as visualization grammars, often claim Wilkinson's Grammar of Graphics as motivation, inspiration or foundation. However, despite the shared intellectual lineage, visualization grammars can vary greatly in design scope and focus. In particular, systems that strongly adhere to domain-specific notions of correctness must trade-off expressivity for validity. This trade-off between semantics and graphical flexibility is currently under-articulated, leading to difficulty integrating various GoG-inspired visualization contributions, especially between InfoVis and Statistical Graphics. We propose three new types of visualization contributions to support such integration: Semantic Validity Theories, Semantic Visualization Systems and Graphical Arrangement Systems. The first two prioritize the adherence of data visualizations to correctness conditions, such as statistical or mathematical transformations and data semantics, whilst the third largely leave validity concerns to the user. We illustrate these new contribution types using examples from two recent grammar of graphics inspired papers, and show how they supported more nuanced notions of expressivity and usability in evaluation and design of visualization systems.
Also Accepted
These submissions were accepted, but their authors are not able to attend in person, and we are unable to support hybrid presentations. They will not be given a talk slot, but their camera-ready versions are included here.
Can the Grammar of Graphics Describe Data Glyphs? Exploring the Application of the Grammar
Abstract
The Grammar of Graphics (GoG) provides a formal and compositional framework for describing how data are transformed into graphical representations. Although grammar-based approaches have been widely applied to conventional visualisations, their applicability to data glyphs has not been fully explored. Data glyphs are composite visual entities that encode multiple data dimensions within a single graphical object and may employ heterogeneous visual elements, mapping strategies, and internal layouts. This work investigates whether the GoG can serve as a foundation for the grammar-based generation of data glyphs. We conduct a theoretical analysis of the GoG layers and apply them to two representative glyphs with distinct structural characteristics: a star glyph and a Chernoff face. The comparison indicates that the star glyph can be described through a relatively direct application of the original layered pipeline. In contrast, the composite structure of the Chernoff face requires component-specific mappings and suggests the need for an iterative application of the graph layers followed by an explicit graphical composition mechanism. These findings indicate that the GoG provides a useful foundation for describing data glyph construction but may require adaptations to support heterogeneous components and complex internal layouts. This work constitutes an initial theoretical investigation and provides directions for future grammar-based approaches to data glyph generation.
Towards a natural language interface for graphic grammars and for specifying visualisation types
Abstract
We propose a system for generating natural language sentences that can be used to describe the declarative, modular and compositional aspects of visualisation grammars, and to specify visualisation types, while being accessible to people who have no acquaintance with formal syntax or coding. The system builds on 'The Languages of Visualisation', which also includes Gestalt aspects of visualisations, and has in turn informed recent systems such as BlueFish and GoFish.