Five months before the conference opened, at the program committee's first meeting, more than twenty names were already written on the whiteboard.
"Are all these names reliable?" the chair asked. The room went quiet for a few seconds—what is the standard for reliable? Fame? Titles? Or who knows the organizing committee?
That day's meeting settled on this improvement:The list can be finalized gradually, but the homework must be done thoroughly first.
The first step in the homework was to create a background card for each candidate. The organizing committee used MuchSci's talent profiles; each time they looked up a person, the four-panel overview was complete:
First panel,Field Alignment—the research fields and keywords in the public profile, and whether they match the conference theme, are clear at a glance. In previous years, there was the awkward situation of a keynote being mismatched with the theme, and the root cause was failing to check this panel.
Second panel,Public CV—education and work experience are presented continuously. The titles on the introduction page have a source draft, and the host's opening remarks can also be written convincingly.
Third panel,Recent Projects—verified project details, with year, type, and funding all included. This is where you check whether the candidate has remained active in this direction recently.
Fourth panel,Affiliations—public roles in societies, journals, and committees.
Three weeks later, the background cards for more than twenty names were on the table. The tone of discussion changed: no longer "I think A is pretty good," but "A's field alignment is a complete match, and A has been active in projects over the past three years, but A's direction overlaps with B's. If we have to choose one..."
With facts as a foundation, disagreements turned into discussable questions.
At the final meeting, the list was finalized. The chair wrote a sentence in the minutes:"Background materials are generated from public data; the final list is decided by the committee."—tools do the homework, people make the call, and this conference's keynote did not go wrong again.