2009Q3 Reports: Publications Chairs

Publication Chairs Report (Barzilay, Regina and Chang, Jing-Shin)

The publication process this year has benefited significantly from the works of previous publication chairs, in particular, the ACLPUB package from Jason Eisner and Philipp Koehn (2005), templates for A4-size proceedings from Olivia Kwong (2006), detailed How-To instructions on publication and ACLPUB from Joakim Nivre and Noah Smith (2008).

We have also made some customized templates for the title pages, copyright pages (with ISBN for each book), citation-stamps, header files and meta files for each book. With these templates, the workshop book chairs should be able to concentrate themselves with the manipulation of submissions with ACLPUB.

Since not all workshop book chairs are familiar with the working environments that ACLPUB assumes: linux/unix/MacOS-based OS, knowledge on basic perl & latex, we have prepared some quick instructions for them to be familiar with the package, some simple instructions on how they can convert their submissions to an aclpub-acceptable form if they are not collecting papers with the START system, and encourage the book chairs to try it earlier.

In very special cases, we simply collect the submissions and compile the proceedings directly. In the end, some workshops with very small number of submissions, including demos, tutorials, and one workshop are compiled by us.

Since the camera-ready day of the workshops, we had close communications with workshop book chairs to resolve various kinds of publication problems, including copyright form issues and technical issues with ACLPUB. We are grateful for workshop book chairs who co-work closely with us to make the publication process smooth. Different working experiences and background with the book chairs however inevitably results in different dates of completion of the proceedings. Therefore, how to promote the aclpub in advance will be an issue to meet the limited time slot available for proceedings compilation.

Like other years, the paper size setting of the source PDF files remains the major problem in compiling the proceedings, since it introduces incorrect margins. Among the main proceedings papers, about 1/4 of the papers are called for re-submission for this problem. We should therefore suggest the START system to detect incompatible paper size in advance. This should save time for re-submission significantly.

In addition to the paper size problem, accent characters in author names, special LaTex symbols (such as '&') and math symbols (such as $epsilon$) in paper titles, inconsistent spelling of names for authors with multiple papers, and author names that have no distinction between first and last name are the special problem for proceedings compilation. We will propose to improve the processing capability of the START/ACLPUB system in these issues to further automate the proceedings compilation process.

Finally, we should probably collect the most commonly encountered error messages from aclpub and solutions, adn compile an FAQ so book chairs of future years can benefit from the real experiences of previous book chairs.