Site experiment · Publication orientation

Paper Constellation

Paper Constellation is a second way to browse my publications. It connects five accepted papers through three questions I keep returning to—Design, Evaluate, and Situate—while future projects stay anonymous. On desktop it is an atlas; on a phone it becomes a chronological trail.

5 accepted papers 3 research threads 7 anonymous future nodes 9 reviewed connections

Explore the constellation Download the reproduction guide

How the map took shape

1. Starting point

   The bibliography had the papers, but not the through-line.

   I could answer “what did you publish?” but not “which question kept following you into the next paper?” The first map put those recurring questions on the page.

2. First mobile attempt

   Squeezing the atlas made the constellation disappear.

   Desktop had room for clusters and paths. On a phone, the labels crowded the sides and long empty year gaps took over the screen.

3. Current version

   Desktop stays an atlas; mobile becomes a trail.

   The chronological vertical thread trail keeps the same papers, memberships, nine edges, and privacy boundary, but gives titles enough room to read.

Three recurring research threads

I placed each paper against three questions I keep asking:

• Design: How can we make useful alternatives and dimensions visible?
• Evaluate: What evidence from people, artifacts, or traces should change the next iteration?
• Situate: How should a representation or assistant fit its physical and practical setting?

Citation counts affect only a quiet halo. These are reading aids, not rankings; position does not encode prestige, venue, author order, or impact.

I reviewed the coordinates, thread memberships, and nine connections by hand. Their positions stay fixed, so DesignWeaver, Physion, and HotSpot do not wander between visits or screenshots.

The ordinary bibliography still loads first and keeps every title, author, venue, and link. The constellation is an extra way to read that same record.

Follow one line: DesignWeaver to What Happened and Why

DesignWeaver helps novices notice and reuse visual dimensions hidden inside prompts and generated concepts. I placed it mainly in Design, with a bridge to Evaluate: surfacing alternatives only helps when people can compare and learn from them.

What Happened and Why? asks how traces of creative work might locate consequential moments, then pair those moments with in-flow explanations from the person doing the work. It moves toward Evaluate while keeping a Design bridge. The line between the papers marks a question I kept: what evidence helps a designer make the next iteration more deliberate? It is not evidence that one paper caused the other.

Try hovering or focusing a paper to see its neighbors. Select it to keep that path open and highlight the same paper in Scholar Lens. Filters travel in the other direction too, so the list and map stay on the same paper.

One record, two reading shapes

The first phone version simply compressed the atlas. Labels fought for the sides and empty years dominated the page. The replacement keeps chronology running downward while three colored rails carry the research threads behind each paper.

On the live page, select a paper to open its details below; press Escape to close it. The mobile trail keeps the same five papers, seven future nodes, and nine connections.

Credit and adaptation

John Thompson shared Nadieh Bremer’s Royal Constellations with me. Bremer’s work at Visual Cinnamon made a useful interaction lesson vivid: stable semantic structure, contextual dimming, and a neighborhood that becomes legible through focus and selection. Her process story also shows how much care sits behind an apparently effortless map.

I carried over those interaction principles, not the royal-family data, visual layout, assets, D3 force-directed SVG implementation, or shortest-path behavior. This version uses server-rendered HTML and fixed SVG paths, keeps the bibliography complete, and gives touch and keyboard readers the same controls.