AVI 2026 Workshop

Interfaces to the Futures of the Past

Workshop on sociable interfaces for human–AI co-production: exploring, improving, linking, and narratively presenting historical data.


Venice. June 8-9 2026.

Workshop overview

This workshop brings together researchers and practitioners interested in how interactive and exploratory interfaces can support human–AI co-production when working with complex historical and cultural-heritage data. It focuses on the design of sociable, participatory visual interfaces that enable users to explore, correct, enrich, and narratively interpret machine-learning–generated outputs, particularly within semantically modelled and linked datasets.

The workshop is discussion-driven and design-oriented: we will use short presentations, collaborative activities, and structured group work to surface shared challenges (scale, uncertainty, bias, interpretation) and identify practical design requirements.

At a glance

Main contact
Andrew Richardson
University of Northumbria
Date
June 8–9, 2026
Location
Venice
Duration
Full day
Themes
Human–AI Co-production Exploratory Visualisation Knowledge Graphs Historical Data Sociable Interfaces

Workshop detail

The use of semantic knowledge graph databases offers exciting possibilities for new ways of modelling historical data, to allow nuanced exploration of lived experience, with broader potential application. Such databases, though, require a significant volume of data to be modelled and for great attention and care to be taken its preparation and linkage. Machine learning tools can contribute to solving the first part of this challenge by applying acquired principles to data at scale but with considerable unreliability and error rates; sustained human engagement can identify and resolve problems of reliability—correcting specific errors while also identifying issues of mis-conceptualisation at a higher level.

Dynamic, interactive or ‘conversational’ forms of data visualisation can, in principle, engage and hold the necessary human attention, including through narrative construction and exploratory curiosity directed towards contributory participation: motivating users to annotate or correct where errors are encountered. Such interfaces often need unfamiliar forms and new visual grammars to be learned. If one accepts that this is desirable, what is good practice to encourage the necessary investment of time and attention?

The workshop opens a conversation around this formulation of a ‘social machine’ and its potential visual interfaces, inviting contributions around themes of interest to the AVI community.

Call for Contributions

We invite researchers and practitioners in HCI/AVI, Data Visualisation, Digital Humanities, Cultural Heritage, and Human–AI interaction to contribute to a collaborative, discussion-driven workshop on sociable interfaces for exploring and improving machine-learning outputs, particularly within semantically modelled and linked historical datasets.

How to contribute

  • Send a brief (>100 word) abstract for your proposed contribution: e.g. position papers, demo presentations or a more general outline of your interest. We will respond to what's submitted, shaping the event around participants specific interests and contributions.

Submissions will be used to curate the program and seed discussion; they are not archival publications.

Where to submit

Submission by email to : andrew.richardson@northumbria.ac.uk

Important dates

Submission Deadline March 29, 2026
Notification of Acceptance April 10, 2026
Workshop Date June 8–9, 2026

Topics

  • User Experience Design for Social Machine Interfaces: How end users engage with needs-specific interfaces for data exploration, validation and improvement (including bias/fairness).
  • Interface Design Principles for Human–AI Co-Production: Principles that encourage agency in identifying and addressing errors or bias in profuse machine-learning processed data.
  • Modular ‘Read–Write’ Design Principles: Modular components of explorable ‘read–write’ data visualisation enabling user-configuration of adaptive interfaces and local “sub-social-machines”.
  • Interfaces for Visual Knowledge Creation and Storytelling: Designing modular interfaces to encourage hypothesis formation using semantic data, and sequential presentation as affectively-engaged ‘visual storytelling’.
  • Thematic Network and Funding Development: Shaping ideas into network activities (UK Research Councils) or design collaborations (EU Horizon programme).

Expected outcomes

The primary expected outcome is a survey of requirements and considerations for the design of a component-based suite of exploratory visualisation tools. The organisers are already leading a small-scale university-funded bid development process in this area; participants may be invited to join follow-on workshops or future international funding applications.

Audience

Although focused on historical data improvement and analysis for academic research and public engagement, the challenges surfaced have wider relevance. The workshop will appeal to those working with rich but complex and “messy” humanistic data—capturing lived experience qualitatively as well as quantitatively— and could equally serve those concerned with society, politics, or culture more broadly.

Format

The opening session will comprise introductory agenda-setting presentations by the leads (30 minutes), two context-setting invited talks (30–40 minutes total, as required in light of submissions), and paper and demo presentations (approximately 4–10; length depending on the number of proposals received, from lightning ‘position papers’ to 15–20 minutes), followed by plenary discussion pre-lunch.

The afternoon includes two participatory group co-design sessions: first to collaboratively set agendas (informed by morning presentations and thematically grouped by facilitators over lunch), second to address them, with self-sorting of groups between sessions. The day concludes with a plenary discussion channelled through a panel of group leaders.

Organisers

Andrew Richardson: University of Northumbria

andrew.richardson@northumbria.ac.uk

Alex Butterworth: Science Museum / University of Sussex

alexbutterworth@alexbutterworth.co.uk

Organisers’ Bios

Andrew Richardson is Assistant Professor in Interaction Design at Northumbria University and a practitioner–researcher specialising in design-led interactive and exploratory data visualisation. His research focuses on developing visual interfaces as research instruments for sense-making, interpretation, and discovery within large, complex datasets, with extensive application in humanities and cultural-heritage contexts. He has co-led the design and development of exploratory visualisation interfaces on major AHRC-funded projects including Tools of Knowledge and Congruence Engine, working closely with historians, curators, and technologists. He is a practice-based researcher with experience organising and facilitating interdisciplinary workshops, design sprints, and research-led demonstrations.

Alex Butterworth is a historian who has worked with digital media for the exploration and communication of the past across a wide range of forms, from games, through locative media, to exploratory data visualisation interfaces, as well as in publications for both trade and academic audiences. In the digital humanities field, he has a specific interest in the use of semantic knowledge graphs and machine learning, and in the challenge of designing interfaces to allow intuitive, iterative hypothesis-building and testing, and the narrative communication of new insights. He has been a Senior Researcher at the Sussex Humanities Lab and the Science Museum, London, and as Co-Investigator has led the digital research on two major AHRC-funded projects (Tools of Knowledge and Congruence Engine — Towards a National Collection). He has participated in data visualisation design sprints for national and international projects including Reassembling the Republic of Letters, for whom he has also led data visualisation workshops.