Designing Better Data Conferences After IASSIST 2017

IASSIST 2017 gathered researchers, librarians, data professionals and educators around a clear proposition: data becomes more useful when people share a common language. Held at the University of Kansas in Lawrence from 23–26 May 2017, the conference connected practical research support with emerging conversations about big data, deep learning, digital agriculture and global food security. Its archived programme remains valuable because it shows how a conference can link technical change with institutional practice.

The event also demonstrated the importance of context. A session on machine learning means something different to a statistician, a research librarian, a farmer, a government analyst or a scholar working with sensitive social data. A successful conference therefore needs to translate between communities without flattening their differences. That lesson is especially relevant in Australia, where research networks span major cities, regional universities, remote communities and government agencies.

For Australian organisers, the archive offers a useful model for balancing ideas with logistics. Delegates may arrive from Sydney, Melbourne, Brisbane, Perth or overseas, often combining conference attendance with teaching, fieldwork or family responsibilities. Reliable public transport information, realistic scheduling across time zones and clear guidance about accessibility can shape the experience as much as the keynote programme.

The strongest lessons from IASSIST 2017 concern design rather than any single technology. Future events should create meaningful opportunities for collaboration, explain data practices in plain language, respect local legal and cultural settings, and preserve useful records after the closing session. These principles can guide both large international meetings and smaller Australian workshops.

Make A Shared Language The Central Design Principle

The phrase “data in the middle” captured a practical challenge: data sits between disciplines, organisations and stages of the research lifecycle. It is collected by one group, cleaned by another, interpreted by a third and reused by audiences who may never meet the original creators. A conference built around this reality should make translation a visible part of its programme.

That means avoiding a schedule in which each discipline speaks only to itself. A technical presentation about neural networks could be paired with a session on metadata, research ethics or community governance. A discussion of digital agriculture could include agronomists, farmers, statisticians, librarians and policymakers. These combinations make the subject accessible without treating specialist knowledge as unimportant.

Terminology deserves deliberate attention. “Open data”, “sharing”, “access” and “reuse” can carry different meanings across universities, public agencies and Indigenous communities. Programme descriptions should define key terms, explain assumed knowledge and identify the practical problem a session addresses. Short glossaries, pre-conference primers and accessible abstracts can help delegates decide which sessions are relevant.

The same principle applies to commercial technologies. A presentation involving analytical data tools should explain the instrument, workflow or service in terms of research decisions and data quality, rather than relying on brand familiarity. This approach helps attendees understand where a technology fits within a broader evidence chain.

Programme For Workflows, Not Just Topics

IASSIST 2017’s range of subjects suggests that future conferences should organise content around the journey of data. Delegates need to see how a dataset is planned, collected, documented, stored, analysed, published and preserved. A topic-based programme can be informative, but a workflow-based programme reveals connections that may otherwise remain hidden.

For example, a stream on digital agriculture might begin with sensor collection and calibration, move into data cleaning and spatial analysis, then examine ownership, farmer consent and long-term access. A deep learning stream could cover training data, model evaluation, reproducibility, bias and communication with non-specialist users. Such sequencing makes the programme useful to attendees at different career stages.

A well-designed schedule should mix formats. Plenary sessions create a shared reference point, while workshops allow delegates to practise skills. Lightning talks make room for early-career researchers and small projects. Birds-of-a-feather meetings support informal exchange, and poster sessions can give participants a lower-pressure way to discuss incomplete or exploratory work.

Useful programme elements include:

• Hands-on sessions using realistic datasets rather than polished demonstrations
• Cross-disciplinary panels with clearly assigned questions
• Lightning talks for emerging projects and early-career researchers
• Clinics on metadata, repositories, licensing, privacy and preservation
• Facilitated discussions that produce a visible record of decisions or resources

Australian conferences can adapt this model to local research conditions. A session on agricultural data may need to address drought, variable connectivity and the needs of regional producers. A workshop involving health or social data should reflect the Privacy Act 1988 and the requirements of institutional ethics review. Presenters should distinguish between general principles and obligations that apply to particular Australian jurisdictions or sectors.

Treat Access And Logistics As Part Of The Scholarship

The practical information preserved by the IASSIST 2017 website is a reminder that the delegate experience begins before the first session. Transport, check-in, venue maps, accommodation and local recommendations reduce uncertainty. When these details are easy to find, attendees have more energy for discussion and collaboration.

Australian organisers should plan for the country’s geography and everyday travel habits. A meeting in Melbourne may be straightforward for local delegates using the train, tram or Myki system, while an event in Brisbane may require clear advice about airport connections, heat and walking distances. Delegates from Perth or regional campuses may need additional time and budget for flights. A conference website should state whether public transport is practical, whether venue access involves stairs, and how far hotels are from the main site.

Scheduling should also respect ordinary working patterns. Early starts can be difficult for delegates who rely on public transport, care for children or have travelled across several time zones. Long lunch breaks can support informal networking, but they should not assume that everyone wants to purchase an expensive meal or consume alcohol. Clear dietary information, water stations, affordable food nearby and spaces for quiet work are simple ways to make participation more equitable.

Accessibility needs to be considered at every stage. Captions, readable slides, live transcripts, step-free routes and an online participation option should be planned rather than added at the last minute. A hybrid stream is useful only when remote delegates can see, hear and contribute. Organisers should publish contact details for access requests and explain how personal information will be handled.

A conference is also part of its host community. Local recommendations can include independent cafés, libraries, galleries and affordable restaurants, rather than presenting the city as a backdrop for visitors alone. In Australia, organisers should account for heat, sun exposure and sudden weather changes, especially when delegates walk between accommodation and venues. These practical details communicate respect for the people attending.

Build Trust Around Data Governance

Technical sophistication does not create trust by itself. Delegates need to know who collected a dataset, who can access it, what consent was obtained, how it may be reused and what happens when an error is discovered. Sessions on governance should sit alongside sessions on artificial intelligence, high-performance computing and visualisation, rather than being treated as administrative side notes.

Australian data conferences have a distinctive responsibility to discuss Indigenous data sovereignty. Research involving Aboriginal and Torres Strait Islander peoples should address authority, community benefit, cultural sensitivity and appropriate control over collection and reuse. These questions cannot be resolved through a generic open-data policy. They require engagement with relevant communities and recognition that valuable data may not be suitable for unrestricted publication.

Privacy is another essential part of the programme. The Privacy Act 1988, Australian Privacy Principles, state and territory requirements, health-sector rules and university policies may affect the same project in different ways. A conference session can help by using practical cases: linked administrative records, geospatial information, interview transcripts, biometric data and commercially supplied datasets. Delegates benefit from hearing how researchers manage uncertainty, not just from being shown idealised compliance models.

Trust also depends on reproducibility. Speakers should disclose data sources, software versions, analytical choices and limitations where possible. If confidential or restricted data prevents full release, presenters can still provide synthetic examples, code documentation, data dictionaries or a clear access pathway. Research librarians and repository specialists are well placed to explain how documentation supports future reuse.

A strong data conference makes room for disagreement. Some communities may favour open publication, while others may prioritise safety, sovereignty or commercial confidentiality. Moderators should establish respectful ground rules and ensure that governance perspectives are not reduced to a final five-minute response. The goal is a more realistic understanding of responsible data practice.

Preserve The Conversation And Measure Its Value

An event’s influence should extend beyond the final plenary. The IASSIST 2017 archive shows the value of retaining schedules, presentation information and practical guidance in a form that remains discoverable. Future organisers should treat the website, repository and conference record as research infrastructure rather than temporary marketing material.

A durable archive can include accessible slides, abstracts, speaker biographies, session recordings, transcripts, posters, workshop materials and links to related datasets. Contributors should be asked for permission in advance, and each item should have clear licensing and attribution information. Files need stable names, descriptive metadata and formats that are likely to remain usable.

Post-event communication should help delegates turn discussion into action. A concise synthesis can identify recurring themes, unresolved questions, useful tools and commitments made during the meeting. Follow-up working groups can report on progress at a later event. This is especially valuable for early-career researchers, who may need a visible pathway from a conference conversation to collaboration, publication or professional development.

Organisers should measure value carefully. Attendance numbers matter, but they reveal little about learning, inclusion or lasting collaboration. Useful evaluation can combine surveys, session feedback, website analytics, repository downloads and follow-up interviews. Questions should test whether delegates gained a skill, found a collaborator, changed a data practice or accessed a resource after the event.

The following framework translates these lessons into practical choices for future events:

Conference priority	Weak approach	Stronger approach	Evidence of success
Shared language	Discipline-specific jargon	Glossaries, primers and mixed panels	Delegates report clearer cross-field understanding
Programme design	Isolated topic sessions	Workflow-based streams and varied formats	Skills are applied after the event
Inclusion	Generic access statement	Captions, quiet spaces, dietary choices and hybrid participation	Broader participation and positive access feedback
Data governance	A single ethics lecture	Governance integrated across technical sessions	Better documentation and more informed debate
Local relevance	Generic tourist information	Transport, cost, weather and community context	Fewer practical barriers for attendees
Archiving	Temporary website pages	Licensed, searchable and maintained records	Continued downloads, citations and reuse

The most durable lesson from the Lawrence conference is that a data event succeeds through relationships among people, practices and infrastructure. Its programme can inspire discussion, but its real value appears when attendees carry methods into laboratories, libraries, farms, public agencies and classrooms.

For Australian organisers, that means designing for distance, diversity and accountability from the start. A conference in Adelaide, Canberra or Hobart may have a different local character from one in Sydney or Melbourne, yet the same principles apply: make the programme navigable, connect technical subjects with social consequences, provide practical support and preserve what participants create. A carefully maintained archive then allows the event to keep teaching long after delegates have returned home.