Area Model for VIS

Introduction

For IEEE VIS 2021 and beyond, the conference changes how it groups shared research interests as part of the reviewing process. Previously, the conference consisted of three sub-conferences, each with similar but separate reviewing processes. Instead, the new area model groups different research topics in Visualization and Visual Analytics into six areas. This allows research papers on closely-related topics to be reviewed in a coherent manner. To ensure a high quality review process, two area paper chairs oversee the reviewing process for each area and draw program committee members from a large joint program committee (PC). The mechanism of a unified PC across all areas allows area co-chairs to have access to wider expertise than was previously the case.

This page provides guidance on how the area model affects authors, reviewers, paper chairs, and the reviewing process more generally. It also gives guidance on how to make an appropriate choice during paper submission.

How Will the Area Model Affect Me?

VIS Areas and Paper Authors

As an author, your main task is to choose an area for your submission. You should think of areas mostly as logistical divisions that ensure a high-quality reviewing process. First, try to find an area for your paper based on the topics grouped in the area. Secondly, if your paper could go into multiple areas, look at the area paper chairs and their expertise. Which area has the area paper chairs with expertise related to your paper? Note that program committee members are not exclusive to an area, so you do not have to worry about identifying which PC members would be suitable to review your manuscript. There are two exceptions to the general recommendations above:

  • Do not submit to an area if you or any of your co-authors are in conflict with both area paper chairs.
  • If you or one of your co-authors is an area paper chair, you cannot submit to your own area.

The FAQ below answers further specific questions about choosing areas.

VIS Areas and Paper Chairs

The area model has two types of paper chairs: Overall Paper Chairs (OPCs) and Area Paper Chairs (APCs). Both take on part of the duties of the paper chairs in the years prior to VIS 2021. Area paper chairs have many of the responsibilities related to the reviewing process: assigning program committee members to submissions, recommending a paper for acceptance, and nominating a paper for awards. The overall paper chairs are there to ensure a coherent reviewing process across all areas, help to resolve conflicts, guide the area paper chairs, and make cross-area decisions, such as best paper awards. Area chairs cannot submit to their own area.

VIS Areas and Reviewers

Each IEEE VIS paper will be assigned two members from the program committee as well as one external reviewer. For external reviewers the reviewing process remains unchanged. Program committee members will see one change: paper bidding will now happen across all papers submitted to IEEE VIS. To ease bidding, mechanisms including updated keywords have been put in place in the conference submission system (PCS) to ensure that it remains manageable.

The VIS Area Model

The use of data visualization can be traced back to at least a millennium ago. The emergence of a variety of graphical plots in the 1800s introduced statistical graphics as a sub-area of statistics. The 1987 NSF report entitled “Visualization in Scientific Computing” prepared for the first IEEE Visualization Conference in 1990. Since then, visualization has become a scientific field, and has expanded to encompass several significant focal points, namely scientific visualization, information visualization and visual analytics as well as many domain-specific areas, such as geo-information visualization, biological data visualization, software visualization, and others. This event series has provided the field of visualization with a prestigious and broad international platform. The early unification effort resulted in the changes of its name to IEEE VisWeek in 2008 and IEEE VIS in 2013. In 2018, the VIS community started the reVISe restructuring process to address the needs for unification and cohesion while maintaining its vibrancy and growth. The area model described here is the result of this reVISe process.

Visualization is the study of the transformation of data to visual representations, which provide cost-effective means for supporting a variety of data intelligence tasks, ranging from rapid observation to in-depth analysis and from model development to information dissemination. Data Science is a new scientific discipline that studies human and machine processes for transforming data to decisions and/or knowledge. Its main goal is to understand the inner workings of different data intelligence processes, such as statistical inference, algorithmic reasoning, data visualization, human thinking, and collaborative decision making, and to provide a scientific foundation to underpin the design, engineering, and optimization of data intelligence workflows composed of human and machine processes. Visualization is one of the core fields in this overarching discipline, and every area in Visualization is thus also an area of Data Science.

Description of VIS Areas

Area 1: Theoretical & Empirical

This area focuses on theoretical and empirical research topics that aim to establish the foundation of visualization as a scientific subject. As such, work submitted to this area may also relate closely to the topics covered in other areas. Work submitted to this area should contribute either theoretical or empirical research.

Theoretical Work

Theoretical work which aims to contribute to fundamental questions that relate to how we understand, assess, categorize, or formalize visualizations and/or visual analytics work. Topics of interest include:

  • Concept Formulation: surveys with organization, synthesis, and reflection; taxonomies and ontologies; guidelines and principles; lexica, syntaxes (grammars), semantics, pragmatics of visualization; and information security, privacy, ethics and professionalism in visualization.
  • Model Development: conceptual models and simulation models for describing aspects of visualization processes (e.g., color perception, knowledge acquisition, collaborative decision making, etc.).
  • Mathematical Formalization: mathematical frameworks, quality metrics, theorems (i.e., mathematically-defined causal relations in visualization).

Empirical Research

Empirical research aims to contribute research methodologies or concrete results of assessments of a visualization / visual analytics contribution or its context of use. Topic of interest include:

  • Research Methodology: general methodologies for conducting visualization research, e.g., typology, grounded theory, empirical studies, design studies, task analysis, user engagement, qualitative and quantitative research, etc.
  • Empirical Studies: controlled (e.g., typical laboratory experiments), semi-controlled (e.g., typical crowdsourcing studies), and uncontrolled studies (e.g., small group discussions, think aloud exercises, field observation, ethnographic studies, etc.), which may be in the forms of qualitative or quantitative *research and which may be further categorized according to their objectives as follows:
  • Empirical Studies for Evaluation: studies for assessing the effectiveness and usability of specific visualization techniques, tools, systems, and workflows, for collecting lessons learned from failures, and for establishing the best practice.
  • Empirical Studies for Observation, Data Acquisition, and Hypothesis Formulation: studies for observing phenomena in visualization processes, stimulating hypothesis formulation, and collecting data to inform computational models and quality metrics.
  • Empirical Studies for Understanding and Theory Validation: studies for understanding the human factors in visualization processes, including perceptual factors (e.g., visual and nonvisual sensory processes, perception, attention, etc.) and cognitive factors (e.g., memory, learning, reasoning, decision-making, problem-solving, knowledge, emotion, etc.)

Example Papers:

Area 2: Applications

This area encompasses all forms of application-focused research, which may aim to solve an application-motivated technical problem, to formulate the best practice in working with domain experts to transform general-purpose visualization technology to domain-specific solutions, to design and develop visualization systems and visual analytics workflows for supporting individual applications, or to gain insight into how to adapt and optimize visualization technology to support the users in a particular application domain. The technical solutions reported in this area are mostly application-specific and usually developed in collaboration with domain experts. These solutions can be in different forms, such as designs of visual representations and interaction techniques, descriptions of algorithms and techniques for data transformation, prototypes of visualization hardware and software, specifications of workflows and best practice, or design studies. Application papers underline the impact and importance of the field beyond the VIS research community itself.

Topics of interest include:

  • Application Domains: In some areas the use has reached a high level of maturity whereas in other domains visualization is emerging as a new and essential component in the workflow. VIS welcomes submissions related to application domains spanning all existing, emerging and potential domains.
  • Application-specific Technical Solutions: visual representations, interaction techniques, algorithms, techniques, hardware prototypes, software prototypes, integrated workflows, recommended working practice, etc.
  • Insight Documentation: success stories and failures about applying visualization technology in practice, achievements of multidisciplinary research projects, benefits gained from collaboration with domain experts, and guidelines resulting from application-focused design studies.

Example Papers:

Area 3: Systems & Rendering

This area focuses on the themes of building systems, algorithms for rendering, and alternate input and output modalities. Papers submitted to this area may present new visualization system architectures, support different computing platforms and development environments, or exploit commodity and specialized hardware devices for either rendering or interaction modalities beyond the desktop. The rendering theme includes algorithms and techniques both in software and through hardware acceleration, and also algorithms for graph layout and label placement.

Topics of interest include:

  • Computing Platforms: commodity hardware, GPU, HPC, energy efficient visualization algorithms and hardware, etc.
  • Visualization Environments: non-immersive and immersive environments, desktop, mobile, web-based, VR/MR/AR, dome theaters, CAVEs, physicalization, remote collaboration, etc.
  • Display Hardware and Output Devices: large and small displays, stereo displays, volumetric displays, 2D/3D printing, non-visual devices, etc.
  • Interaction Modalities: touch, pen, speech, gesture, haptics, etc.
  • Development Environments: programming languages, software libraries, authoring systems, visualization toolkits, software frameworks for integration and interoperability, etc.
  • Processing Paradigms: parallel, distributed, out-of-core, progressive, streaming, in situ, in transit, etc.
  • Engineering Visualization Systems: visualization system lifecycle, testing, performance analysis, verification, validation, etc.
  • Visualization Systems: general-purpose and application-specific plug-ins, apps, tools, systems, multi-system workflows, etc.
  • Data and Software Resources: open data, open source software, benchmark data, reproducibility, authentication, etc.
  • Rendering Techniques: surface rendering, volume rendering, point-cloud rendering, line-cloud rendering, global illumination, stylized rendering, transfer functions, etc.
  • Lighting and Shading Models: volume rendering integrals, spectral rendering, learning lighting and shading models from real-world data.
  • Placement Techniques: object placement, graph layout, word/tag cloud, etc.
  • Other Synthesis Techniques: fabrication, sonification, haptic feedback, etc.

Example Papers:

Area 4: Representations & Interaction

This area focuses on the design of visual representations and interaction techniques for different types of data, users, and visualization tasks. In principle, the data concerned can be of any data types, such as spatial or non-spatial; continuous or discrete; statistic, temporal or streaming; numerical, textual or imagery, etc. The user concerned can be from any user groups (e.g., scientists, scholars, students, analysts, administrators, or the general public) and of any level of visualization literacy and skills. The tasks concerned can be of any operational needs, such as effective information dissemination, rapid data observation, and explorative information seeking. The visual representations concerned can be of elementary encoding (e.g., visual channels, statistical graphics) as well as complex visual mapping (e.g., spatiotemporal data visualization and coordinated multiple views), and can be in visual as well as non-visual forms for enabling visualization via different human sensory devices. The interaction techniques can be based on traditional WIMP (windows, icons, menus, and pointers) and direct manipulation. Papers submitted to this area are normally expected to emphasize their novel contributions in terms of the design of visual representations and interaction techniques, while the work may also discuss the related hardware and software components for data transformation, image synthesis and displays, interaction, and immersion (see also Areas 3 and 5).

Topics of interests include:

  • Visual Channels: geometric channels (e.g., location, size, orientation, shape, etc.), optical channels (e.g., color, opacity, shading, motion, etc.), topological and relational channels (e.g., connection, overlapping, etc.), and semantic channels (e.g., number, text, glyph, etc.).
  • Visual Representations: for textual data, tabular data, relational data (e.g., hierarchy, tree, set, graph/network), geospatial data, temporal data, imagery data, geometric data (mesh-, point-, line-, curve-based data), field-based data (e.g., volumetric, vector, and tensor field), corpus data, multi-type data, uncertain and missing data, models, functions, and procedures (e.g., algorithms and software), etc. in raw, filtered, or transformed (e.g., aggregated) form.
  • Interaction Techniques: UI design for visualization, zoom and navigation, magic lens, query-based exploration, direct manipulation, interactive deformation,natural interaction, user-adaptive interaction, interoperation between interaction and visualization tasks, editing tools, collaborative visualization, etc.
  • Visual Communication Techniques: focus+context design, illustrative and explanatory visualization, stylized visual representations, storytelling and narrative visualization, textual annotation for visualization, etc.
  • Intelligent Visualization and Interaction: Automated visualization generation, mixed-initiative visual interaction, learning UI models for automated capabilities in visualization systems.
  • Technical Discourses on Visual Representations and Interaction Techniques: visual and interactional metaphors, scalability of visual mapping, and interaction costs, 2D vs. 3D representations, static vs. animated representations, visualization literacy, etc.

Example Papers:

Area 5: Data Transformations

This area focuses on the algorithms and techniques that transform data from one form to another to enable effective and efficient visual mapping as required by the intended visual representations. In principle, the source and destination data can be of any data types, such as spatial or non-spatial; continue or discrete; statistical, temporal or streaming; numerical, textual or imagery, etc. Such data transformation, which may sometimes be referred to as wrangling or munging in some other fields, may include extracting information from the source data (e.g., surface extraction from volume data, and network construction from textual data), integrating data from different sources (e.g., multi-modality registration), reorganizing data for efficient processing (e.g., hierarchical data representations), enriching data with additional information and functions (e.g., uncertainty analysis and label generation), and improving data quality and usability (e.g., data cleansing). Papers submitted to this area are normally expected to emphasize their novel contributions in terms of the algorithms and techniques for data transformation, while the work may also discuss the intended visual representations and their generation (see also Areas 3 and 4).

Topics of of interests include:

  • Information Extraction and Data Abstraction: keyword extraction, metadata extraction, surface extraction, feature extraction, pattern recognition, structural and semantic analysis, skeletonization, spatial abstraction, topological abstraction, temporal feature tracking, multi-material interfaces, etc.
  • Data Integration: multi-modality, multi-stage, and multi-level data registration, spatial and non-spatial data integration, multi-field representations, etc.
  • Data Reorganization: voxelization, triangularization, multi-resolution sampling and representations (e.g., discrete sampling, volumetric lattices, wavelet representations), spatial partitioning (e.g., octree, k-d tree, bounding volume), data segmentation, compressed data representations, frequency-domain representations, databases for query-based visualization, etc.
  • Data Enrichment: uncertainty analysis, deformable models, label generation, spatialization, etc.
  • Data Wrangling and Improvement: data wrangling, data re-shaping, data cleaning, data editing, data smoothing, and data modelling.
  • Mathematical Frameworks for Data Transformation: numerical analysis, computational geometry, topological analysis, graph theory, statistical analysis, probability theory, information theory, dimensionality reduction, etc.
  • Machine Learning for Data Transformation: automated discovery of data models and data transformation algorithms for visualization, learning-based parameter optimization of data models and data transformation algorithms for visualization, etc.
  • Technical Discourses on Data Processing and Management in Visualization: feature specification, data provenance, processing provenance, interactive processing, data synthesis, quality assurance, etc.

Example Papers:

Area 6: Analytics & Decisions

This area focuses on the design and optimization of integrated workflows for visual data analysis, knowledge discovery, decision support, machine learning, and other data intelligence tasks. It typically addresses technical problems that cannot be solved using solely machine-centric processes (e.g., statistics and algorithms) or solely human-centric processes (e.g., visualization and interaction). It may also address the need for using interactive visualization to improve the trust, interpretability, understanding of machine-centric processes and their underlying models and need for data intelligence workflows to benefit from theoretical models and empirical findings in cognition, e.g., in areas such as distributed, embodied, and enactive cognition. Hence papers submitted to this area are normally expected to feature an integrated approach. This area includes visualizations and visual analytic tools that show these models, or leverage them heavily for producing the visualization.

Topics of interest include:

  • Integrated Workflows for Information Seeking, Knowledge Discovery, and Decision Making: Typical technical problems may include information retrieval, multivariate and semantic search; classification, pattern recognition and clustering; similarity, correlation and causality analysis; spatiotemporal tracking and movement analysis; event and sequence analysis; multimedia data analysis; anomaly and change detection; relationship, association, hierarchy, network and structure analysis; intention and behavior analysis; factor analysis and dimensionality reduction; uncertainty and risk analysis; and so on.
  • Integrated Workflows for Machine Learning: Typical technical problems may include cleaning and labelling training data; assisting active learning or other semi-automated learning methods; facilitating model testing, evaluation and model comparison; supporting the analysis of learned models and learning processes; enabling model understanding, explanation, refinement, and steering; and monitoring the deployment of machine-learned models as well as other machine-centric processes.
  • Workflow Optimization: techniques, design patterns, and best practices for designing, developing, evaluating, and improving integrated data intelligence workflows. Methods for analysing and alleviating data biases, machine biases, and human biases.
  • Knowledge-assisted Workflows: knowledge acquisition, mixed-initiative workflows, real-time guidance and recommendation, provenance management and utilization, post-action review, knowledge sharing, and analyst training in visual data analysis.

Example Papers:

Frequently Asked Questions

  1. My paper fits equally well into two areas – which one should I pick?
    Pick the area where the area paper chairs are likely the most knowledgeable about your paper. Note that program committee members are not specific to the area and can be chosen by any area paper chair.
  2. What happens if I pick the “wrong” area?
    Due to the unified program committee, it is unlikely that your choice would have a strong negative effect on the choice of reviewers and the quality of reviews for your paper. If the area paper chairs feel strongly that your paper is in the wrong area (or both are conflicted with your paper), they can, in exceptional cases, liaise with the chairs of other areas to propose a move to the authors.
  3. Can I tell who reviews in which area?
    No, the program committee is unified, and PC members are likely to review papers from multiple areas.
  4. My submission does not fit into any area – what should I do?
    Closely review the descriptions of the areas. We expect that all papers within the scope of VIS will be able to find one or multiple areas that are suitable to handle a submission. If you believe that your manuscript is in scope of VIS, but does not fit into any area, please contact the overall paper chairs.
  5. How do areas and keywords relate?
    Keywords are an important instrument to match your submitted manuscript to qualified program committee members, who will indicate interest in papers through a bidding process, be assigned by the area paper chairs to papers that are a good match, then invite competent external reviewers, and who will also write a review themselves. Areas are administrative divisions, what matters most is that you will have area paper chairs that can correctly identify the qualified program committee members and make informed decisions about your manuscript.
  6. How do paper awards work in the new area model?
    Best paper awards and honorable mentions are given out for a certain percentage of all papers. Area paper chairs will nominate papers from their area for these awards. Final decisions on awards will be made independent of areas by a separate committee.
  7. What happens if my co-authors or I are in conflict with both area paper chairs?
    In case of a conflict of interest with both area paper chairs, you need to send your paper to another area. If you are only in conflict with one of them, that can be handled, and it’s fine to submit. To judge whether you are in a conflict of interest refer to the VGTC ethics guidelines for reviewers - which equally apply here. You may also check with the area paper chairs to verify your conflict of interest in case of doubt.
  8. How will areas be reflected in the IEEE VIS program?
    Areas are only an administrative division in the reviewing process. Areas will not be reflected anywhere in the conference program. For example, sessions will be curated independent of the areas papers were submitted to.
  9. How do VIS areas differ from areas/subcommittees in other conferences?
    Compared to ACM CHI’s subcommittees, IEEE VIS has no area-specific program committee. Also, at VIS acceptance is recommended by the area paper chairs and confirmed by the overall paper chairs. At CHI, the area-specific program committee makes a joint decision in the program committee meeting.
  10. Some areas don’t seem as coherent as others. Is that true?
    For some areas, the main subject has enjoyed more structured development over the years and has established a set of relatively well-defined topics. In other cases, an area encompasses a few topics that could potentially become separate areas in the future when the research activities on these topics have reached a certain scale. The area model is expected to evolve in response to the emergence of new topics and the rapid development of some existing topics. This change will be managed by a dedicated area curation committee (review the restructuring proposal and future governance documentation for details).
  11. Which areas will have the best acceptance rates?
    The final acceptance decisions are coordinated by the OPCs with an eye towards a balanced program of consistent quality. There may be variations in acceptance rates across areas, and across years. Variation is a natural feature of a dynamic and vibrant process. Authors should decide which area is most suitable for their paper based on intellectual fit and scope; past numbers may not predict future patterns.
  12. Whom should I contact if I have a question about a specific area?
    Please contact the area chairs of that specific area.
  13. Why are APCs banned from submitting to their own area?
    It may appear unusual that APCs cannot submit to their own area (“self-submission”), given that strong expertise in exactly their area is one of the reasons they are appointed. reVISe considered many alternatives, but ultimately proposed this model for the following reasons:
    • Without self-submissions, avoiding conflicts of interest and leaks of information about the review process at the chair level becomes a realistic goal. As APCs do not have access to information about other areas’ papers, they cannot inadvertently learn about the reviews and reviewers of their own papers, and are completely removed from the decision process happening in other areas that involve their own papers. Avoiding conflicts of interest and potential sources for deanonymization increase the trust that authors have in the review process overall.
    • The unified program committee allows the selection of expert reviewers, supported by the new keyword mechanisms, independently of the area a paper is submitted in. Furthermore, in the analysis of the area model during its development, it became apparent that many papers fit well in several areas. Hence, for the vast majority of cases, not allowing self-submission will not compromise the quality of reviewing, and avoids more drastic restrictions, such as disallowing APC submissions completely.
    • In some cases, an APC paper will not fit well into another area. However, experience with journal editors shows that a managing editor does not have to be an expert in an area, as long as they can rely on a pool of a qualified reviewers, which is ensured by the unified PC. The choice to block APCs from submitting was balanced against other considered solutions such as involving additional shadow paper chairs or chairs from other areas to help out in cases of conflict. The current solution is the one that involves the least chances to reveal anonymous information and the one that is administratively the most simple solution. Hence, the approach taken balances potential unpleasantness with process simplicity and transparency.
    • Finally, the choice to block APCs from submitting to their own area is consistent with practices in other conferences. Paper chair positions are considered an honor taken on by senior community members. Many who fill these roles are in a position in their scientific careers at which they can give priority to service to the community, yet they still are actively involved in scientific research and publication. It is not unusual in other communities to entirely disallow submissions of chairs overseeing a papers process (for example, all SIGPLAN conferences, including POPL, and PLDI; theory conferences such as STOC, FOCS, SODA, SOCG, ICALP, ESA). The model at VIS offers a compromise between striving for reviewing quality and integrity and allowing APCs to still contribute to the scientific content of the conference.

As all other aspects of the area model, this will be closely watched by ACC, and alternatives will be considered if the need arises.

Version History of the Area Model