Invited talks are quite a bit different than refereed papers, and the process used to accept them is also quite a bit different. Two members of the program committee are specifically designated at the IT co-ordinators. This year we are very fortunate to have two excellent folks as co-ordinators. Last year’s conference chair, David Blank-Edelman along with Doug Hughes headed up the effort this year to select two tracks of invited talks.
The original intent of the invited talks track was that the committee would invite prominent individuals to speak at the conference. That’s why they’re called “invited”. But as the conference expanded it was decided that we would be served well by actually soliciting proposals from those who were interested in speaking. We still invite speakers, but we also accept proposals. In order to keep everything fair, we ask those we invite to submit a proposal. This allows us to review all the possible talks in the same manner.
The proposal is not that complicated. We ask for a speaker’s biography, a description of the talk, and an outline of what will be presented. With these proposals the committee needs to choose enough talks to fill two full tracks of 10 slots each, for a total of 20 presentations. Sometimes this number will be a little less. If the committee finds a particularly strong or important presentation that we think everyone needs to year, we will create a plenary session on Thursday.
When choosing talks, we look at several characteristics. We want the subject matter to be relevant. A talk that explains the proof of Fermat’s Last Theorem would be interesting, but hardly relevant. We also want variety among the presentations. Although Spam prevention is important, we don’t want four separate presentations about it. We want to ensure that the presentation is thorough and backed by sound facts and research. Finally, we look at the quality of the speaker.
You can see how much this differs from selecting papers. A paper is selected solely on its own merit, regardless of the authors or their abilities as speakers. The focus is on the research presented in the paper: the presentation is secondary. But an invited talk lasts nearly 90 minutes, and the presentation is the final product. So it is appropriate to consider the presenter’s abilities as a public speaker. But bear in mind that it isn’t the only criteria. Committees are willing to consider average or below average speakers if their topic is particularly important or relevant. And the converse is also true: we would consider a weak topic if it was presented by a strong speaker.
After the submission deadline had passed and we were sure all the lingering proposals had been submitted, the two co-ordinators got together on a conference call and discussed the proposals, generally classifying each as “definitely”, “maybe”, and “no way”. This provides an idea of how many “maybe” talks need to be accepted to fill out the slate. These choices are made based on the criteria listed above: what talks look good and relevant, what provides the best variety, and which speakers appear to be strong. The end result is a list of those talks we want to accept and a short list of alternates. This list was presented to me for final approval. There were a few choices that the co-ordinators were not able to reach a concensus on, so I ended up making the final decision. This year I feel very fortunate that we were able to get a strong slate of talks, due to the tireless efforts of our fantastic IT co-ordinators.
Once the choices were finalized, we contacted the speakers to make sure that they were still interested in presenting. If anyone had backed out, we had alternates to fall back on. This year, it wasn’t necessary to use any of our alternates.
One interesting problem we faced this year was a presentation proposed by one of the most prominent researchers in our community on a topic that the committee felt was particularly important and relevant. However, the speaker told us that he didn’t think he could fill more than 45 minutes. That makes for a very short presentation unless we could find something to pair it with. Unfortunately, we couldn’t, and we ended up in a difficult position. We really wanted to accept this talk, but there was no place on the program to put it. We don’t really have a position for a 45 minute presentation unless we invent one. So that is eventually what we did: we turned it in to the Friday lunchtime presentation, where Dr. Alva Couch will be telling all of us how autonomics will change our industry. I will post more about that at a later date.