2014Q3 Reports: General Chair

2014Q3 Reports: General Chair

Daniel Marcu

ACL’2014 At A Glance

• ACL’2014 received 1,123 submissions, of which it accepted 146 long papers and 129 short papers for a combined acceptance rate of 26.1%. After adding 19 TACL papers, the program was structured to include 159 oral and 145 poster presentations.
• Workshop proposals were handled jointly by ACL, EACL, and EMNLP. Across the three conferences, there were 42 submissions. Based on the workshop organizers’ preferences and the quality of the submissions, ACL’2014 eventually selected 17 workshops (15 one-day workshops and 2 two-day workshops). The ACL’2014 Workshop Chairs have also collaborated with the CoNNL’2014 organizers to make CoNNL’2014 a collocated event.
• ACL’2014 received 19 tutorial proposals, of which it selected 8; 39 demo proposals, of which it selected 21; 7 thesis and 19 research paper submissions for the Student Research Workshop, of which it accepted 5 and 8, respectively.
• ACL’2014 will have two distinguished invited speakers: Corinna Cortes (head of Google Research New York) will talk about her work on learning ensembles of structured prediction rules. Zoran Popović (Professor, University of Washington) will show how humans and machines can collaborate to solve hard problems that neither can solve alone.
• ACL’2014 has secured an impressive number of sponsorships: Baidu (Platinum Sponsor); Bloomberg, Google, Microsoft, Nuance, and Yahoo Labs (Gold Sponsors); Information Sciences Institute and Xerox Research Center Europe (Silver Sponsors); Brandeis University, Facebook, and Yandex (Bronze Sponsors); and IBM Research and the University of Washington (Supporters).

The number of submissions and accepted presentations is consistent with numbers observed in previous years while the sponsorship is higher than in previous years. All indicators point to a field that is striving while continuing to mature.

The organizing committee was outstanding: the chairs of each individual area have acted with high professionalism while fearlessly validating and testing new concepts and ideas. This has resulted in many innovations, from local organization to program structure, to publication approach. Our intention is to collect feedback from the conference participants in order to provide data grounded recommendations for future conferences.

We could have not got this far without the continuous guidance of the ACL Executive Committee, in particular Dragomir Radev and Priscilla Rasmussen, who bring to the table tremendous amounts of institutional knowledge and know-how. We are extremely grateful to them and everyone else who helped make ACL’2014 possible.

The rest of this report highlights only the main innovations and recommendations for future conference organizers. The reports of the individual chairs provide additional details, from challenges to recommendations, to innovations.

Main Innovations

• We scheduled two large poster sessions on two evenings of the conference, to accommodate the large number of poster presentations and give participants more time to socialize.
• Instead of a paid banquet, we will have a no-fee light social event. We hope this will provide a better avenue for socializing and sharing ideas.
• The president’s talk will be given on the first morning of the conference. Since we choose our presidents based on their prominence in the field, we thought it would be wise to give them the opportunity to choose between a humorous presentation, typical of our tradition, a scholar presentation, or a combination of both.
• We optimized the conference schedule based on feedback received from more than 300 conference participants. The response to our solicitation was tremendous.
• We solicited nominations for outstanding reviewers and acknowledged them in the proceedings (around 14% of reviewers were acknowledged).
• Created an XML version of the majority of the papers accepted for publication. This will enable machine-based access to our work and provide the foundation for better research on our own research.

Recommendations

Policies:

• Clarify policy for self-published papers.
• Enable persistence of reviewers and their assignment completion date in START.
• Make explicit what chairs own the synchronization of the conference schedule with the papers associated with the schedule: the publication, program, or local chair?
• Relax the copyright policy, which currently provisions for the transfer of the entire copyright away from the authors (see the recommendations made by the Publication Chairs).

Processes:

• Ensure early creation within START of initial websites for the workshops with their admin users and basic information.
• Expand Student Workshop to undergraduate students we want to attract to do graduate work in our field.

Tools:

• Collaborate with START to integrate Mark Dredze’s reviewer management tool into START.
• Create additional scripts for streamlining the reviewing/publication process (checking non-anonymity and formatting; enabling merging of schedules and metadata; etc.)