A tour of the data visualization gallery at IASSIST 2017

The archived IASSIST 2017 programme turns the University of Kansas into a digital wayfinding site for anyone who could not join the event in Lawrence on 23–26 May 2017. Its central theme, “Data in the Middle: The Common Language of Research,” suits a gallery tour because each presentation acts as a doorway into a different way of seeing evidence. Big data, deep learning, digital agriculture and global food security are not isolated subjects here; they meet in a shared question: how can complex information become clear enough to guide people? The practical pages for check-in, transport and local recommendations extend that idea, showing how conference information can meet a visitor at the point of need.

A useful tour begins by treating the archive as a thematic exhibition rather than a literal set of rooms. The original gathering has passed, but its programme still preserves an intellectual sequence: what was measured, how it was modelled, which decisions it might support and what audiences need in order to trust it. From Australia, this is especially valuable because research communication crosses very different scales, from a paddock and a regional service town to national agencies and global food chains. The gallery offers a chance to compare those scales without pretending that one visual form suits every setting.

Gallery theme	Core evidence	Visual forms to notice	Main interpretive prompt
Big data	Large, varied and continuously changing records	Maps, networks, time series and filtered dashboards	Which patterns remain meaningful at scale?
Deep learning	Training examples, learned features and model behaviour	Layer diagrams, heat maps and confidence overlays	Where can uncertainty still be seen?
Digital agriculture	Field, machinery, climate and market observations	Satellite mosaics, sensor traces and yield maps	How does a signal become a farm decision?
Global food security	Production, access, trade, policy and nutrition	Flow maps, comparative charts and timelines	Which causes and consequences are connected?

Building a route through the archive

The first stop is the conference frame itself. “Data in the Middle” suggests that data is neither a decorative supplement nor a final authority. It sits between observation and action, requiring researchers, designers and communities to interpret it. That position becomes visible in the plenary sessions, where broad questions frame the more focused contributed presentations. Following that structure gives visitors a coherent route through the archive: begin with the purpose of the evidence, move toward the means of analysis, and finish with the people expected to use the result.

The archived schedule adds another layer. Instead of scrolling through an undifferentiated list of papers, visitors can follow sessions chronologically and see how ideas developed over the four days. A talk about a data-intensive system may sit beside one examining communication between specialists, while discussions of food, farming and machine learning reveal the interdisciplinary character of the event. The result resembles a set of connected rooms, even though each destination retains its own subject and speaker.

Attention should also shift from the image on the screen to its surrounding material. Captions, definitions, units and methodological notes determine whether a visual explanation can be understood outside the original room. Colour carries meaning only when the legend makes that meaning recoverable. Scale requires context, and a dramatic axis can shape interpretation more strongly than the underlying numbers. In a well-constructed gallery, those features become part of the exhibit rather than administrative detail.

Big data and deep learning made visible

Big data displays often promise an expansive view: dashboards crowded with indicators, maps covered by layers, or networks whose dense connections initially appear to promise universal insight. The more revealing question is how the display moves from volume to meaning. Which filters have been applied? Which changes in the data are important? Does the design distinguish a genuine shift from a reporting artefact? The strongest visualisations make those decisions visible rather than allowing a polished chart to conceal its own construction.

Deep learning adds another layer of opacity. A layered diagram can explain the broad architecture of a model, but it cannot by itself reveal why a particular output occurred. Confidence overlays, highlighted features and comparisons across examples help, although they should not be confused with complete explanations. In the gallery, model performance and model interpretation remain separate stops. One asks how accurately a system responds; the other asks whether a person can understand what shaped that response. Clear labels help audiences hold both questions at once.

For Australian viewers, scale is an especially immediate concern. Distances are expressed in kilometres, settlements can be separated by vast expanses of land, and a narrow change in rainfall can affect widely separated communities. Maps of heat, water or population density therefore carry familiar implications here. So do dashboards about drought and bushfire risk, where visual clarity can influence public trust. The geographic challenge also raises questions about missing data, especially where remote communities have weaker connectivity or fewer recorded observations.

Digital agriculture and global food security

Digital agriculture brings the gallery’s separate strands together. A yield map may combine satellite imagery, soil sensors, machinery records and weather observations, but the farm decision ultimately depends on more than a coloured pixel. Does the map distinguish a profitable management change from a visually attractive pattern? Can the farmer verify it against local knowledge? Is the equipment compatible, the service reachable and the advice affordable? The visual narrative works best when it follows the chain from measurement to action without removing the people who operate that chain.

Global food security broadens the scale again. Maps of production and trade can reveal dependencies between regions, while comparative charts show how food availability and nutrition outcomes differ. They can also flatten a complex system into a single metric. A country may export large quantities of food while households within it experience insecurity. A low-price commodity may support consumers but place pressure on producers. A responsible gallery display links cause and consequence, identifies the period and unit behind each number, and leaves room for lived experience.

The Australian market context adds useful friction to that discussion. Agritech companies often operate through a mix of university research, government programmes, producer organisations and private investment, with contracts and forecasts commonly expressed in Australian dollars. Producers may work across domestic supply chains and export markets exposed to exchange-rate swings. Even local language carries assumptions: meeting “at three in the arvo” means the afternoon, while a stop at the “servo” may be part of a longer route between towns. These practical details belong beside global statistics because they shape who can use data, when it arrives and what action it permits.

Viewing the gallery from Australia

Australian audiences bring useful visual habits to a research gallery. They may be accustomed to sporting statistics presented through positions, possession, speed and marginal percentages, but they are equally alert to the danger of letting precision imply certainty. Dense dashboards should not overwhelm a simple finding, and animated presentations should not conceal the time required to interpret them. The most accessible exhibits allow a quick first reading followed by a more detailed path for interested visitors.

Canberra’s role as a home of national policy and research organisations makes it a useful reference point. A presentation on agricultural data can be considered alongside how public agencies define evidence, coordinate across regions and communicate risk. Melbourne adds a different urban perspective: data teams and creative communities often gather around compact laneway venues, where the social setting is informal but the work can be highly technical. Ordering a flat white between sessions may not advance a research project, yet these details illustrate how collaboration develops around specialised knowledge in Australian cities.

For a sports comparison, the championship visual story offers a familiar Australian parallel. Competition information becomes persuasive when movement, position and audience context are shown together rather than reduced to a leaderboard. Sydney’s familiarity with several sporting codes makes that translation especially easy, but the underlying visual discipline travels well: spectators need to know what happened, why it mattered and where uncertainty lies. The same approach can guide a chart about crop yields, rainfall or public health.

Planning a visit and reusing the archive

The conference archive preserves practical information as well as research. Visitors can use the check-in guidance, schedule and transport notes to understand how the event was intended to work around the University of Kansas campus and Lawrence. Since this is an historical resource rather than a live event listing, date-specific instructions should be read as a record of the 2017 arrangements rather than current travel advice. That distinction is especially important for international readers, whose expectations of registration desks, campus access and local travel may differ.

Local recommendations offer another way to understand the conference setting. A walk through Lawrence, a meal near campus and the time allowed between sessions can influence which presentations people attend and which conversations continue afterwards. Australian travellers should adjust for familiar differences such as distances measured in miles rather than kilometres, weather that can shift quickly, and tipping customs that differ from those at home. These details are small beside the programme, yet they reveal how access to knowledge depends on more than technical availability.

Reusing the archive can extend its value well beyond the conference. Educators can build a visual-literacy exercise around a single chart, while researchers can trace how agricultural evidence connects with machine learning or food-security policy. A digital exhibition assembled from the presentations would also show how captions, sources and visual hierarchy affect understanding. What survives is a method of seeing: assemble evidence carefully, reveal its shape, and preserve enough context for the people who must act on it.