CHI

Welcome to CHI 2009! CHI comprises many events, ranging from archival material stored in the ACM digital library, to transient interactions such as the presentations, panels, and poster discussions, to the many social interactions and activities that make CHI a collegial and intimate experience. All parts are important, but it is the archival material -- especially the papers and notes -- that establishes CHI as the leading academic conference in Human Computer Interaction. Yet there are significant challenges in managing paper and notes within CHI. The HCI field has been very successful at creating new generations of research and practitioners over the years. The many people who are part of this community see CHI as the place to share their knowledge and experiences with others, primarily by publishing and presenting papers and notes. This has stressed the system in many ways. As submissions increase, so do the difficulties in managing the review process, finding good reviewers and other volunteers, matching papers to those competent in the subject matter, deciding which papers to accept or reject, maintaining consistent standards across both paper and notes, and not falling into the trap of overly narrowing our view of what is an 'acceptable' CHI paper.

This year, we introduced several large changes to the CHI papers/notes process to mitigate these challenges, most which will be transparent to attendees. First, we reorganized the CHI program committee into nine topical subcommittees - each a mini program committee - comprising sub-committee chairs (SCs) and various associate chairs (ACs) knowledgeable on the topic. Authors could select the subcommittee that he or she felt could best handle their submission. We did this to improve the match of a submission to AC and ultimately to reviewers, to have more focused and relevant discussions in the program committee meeting, and to minimize the load on individual volunteers. Second, we combined papers and notes, where all were handled in exactly the same way. We did this to ensure a consistent decision standard across both submission types. Third, we introduced contribution types, where each type described a different way that a CHI submission could contribute to the field as well as typical questions such a contribution should address. Authors identified their submission by contribution type, and (hopefully) used the information to help structure their paper. The idea is that we wanted to encourage a broad variety of submissions from authors (rather than 'formula' papers), while also providing guidance to referees by supplying criteria appropriate to the type of contribution the submission was making.

It will likely take several years before the full impact of these changes are known. We know that subcommittees did help us manage the large number of submissions. We also believe there was an overall better match between referees and submissions, and that papers and notes were handled consistently. We don't yet know about the effect of contribution types: this is a cultural change where we are hoping that authors will be more willing to write papers that don't match a particular formula, and that reviewers will be more accepting of those submissions.

Now for the numbers. This year, there were 1130 submissions, comprising 711 full papers and 419 notes. This is the highest number of submissions ever to CHI. Of these, we accepted 24.5%. The papers/notes committee involved 107 volunteers: the 2 co-chairs, 10 sub-committee chairs, and 95 associate chairs (ACs). Each AC managed 10-14 submissions, and personally recruited at least three -- sometimes more -- referees knowledgeable in the paper's topic. Refereeing was through blind review. Each referee returned a recommendation along with a detailed review, and authors had opportunity to rebut these reviews. Additional reviews were sometimes solicited. Almost all program committee members then attended a two day meeting in Boston in December. Rigorous discussions took place at the PC meeting, and the majority of papers were read by a second AC as well. The decision process was highly visible so that the committee could calibrate itself.

Finally, the various committees nominated 5% of the submissions as potential best papers. A separate committee deliberated over these papers, where only 1% of papers and notes received a best paper award. In total, as you will see in the program, 32 papers and four notes were designated as honorable mentions, while seven papers and four notes honored as best papers. Congratulations to all authors who achieved this significant status!