2011Q3 Reports: Student Research Workshop Chairs

From AdminWiki
Jump to: navigation, search

This year the Student Research Workshop was renamed the Student Session. This was done in an effort to rebrand the meeting so to increase the number of submissions, and generate greater interest to the general research community.

Program Committee Formation: The co-chairs of the ACL 2011 Student Session are Sasa Petrovic (University of Edinburgh, UK), Emily Pitler (University of Pennsylvania, USA), and Ethan Selfridge (Oregon Health and Science University, USA). The faculty advisors are Miles Osborne (University of Edinburgh, UK) and Thamar Solorio (University of Alabama at Birmingham, USA). The program committee list was approved by the Student Session faculty advisors on September 10, 2010. The program committee consists of 55 members, of which 31 are from North America, 20 are from Europe, and 4 are from Asia/Pacific. They represent 25 areas of NLP andcomputational linguistics research.

Conference Setup and Procedures: The call for papers was first uploaded to the ACL 2011 website on August 24, 2010. We also emailed it to a number of lists on Monday September 20th, 2010. This year we put more effort in attracting submissions from bachelor and master-level students in an attempt to get students interested in computational linguistics earlier in their careers. This was reflected in the call for papers. A START account was created for the submission process.

Submissions and Reviewing Procedure: We introduced a "Thesis Proposal" submission type, in addition to the traditional "Research Paper". 57 valid submissions were received. Of these, 9 were Thesis Proposals and the rest were Research Papers. All of the submissions were assigned 2 reviewers. We accepted 22 papers total, of which 4 were thesis proposals, and 18 were research papers. Thus, the acceptance rate was 38.6%. We believe that the addition of Thesis Proposals increased the number of submissions since, in all likelihood, the research work by many of the authors would be better suited for the main conference.

Presentation format: This year, all of the papers will be presented as posters during the main poster session of ACL conference on June 20th. We decided on this format to avoid clashing with other tracks in main conference, which has resulted in poor attendance of the student session in the previous years.

Paper Mentors: Each accepted paper received a mentor, a senior member of the research community, who will meet with the student author during the conference and provide feedback. The mentor will have read the paper and provide comments on both the general research direction and on specific details related to the paper.

Funding for Participants: We have been awarded funding for participants from the ACL Walker Student Fund, the European Association of Computational Linguistics, Google, and the National Science Foundation (NSF Award Number 1102435). Each accepted Student Session paper received free main conference registration ($185), $300 towards accommodation, and a portion of their airfare subsidized. Despite the generous grants, we were still very short on money. This was especially true for authors outside the U.S who were not eligible for NSF support. The total airfare alone was $23,331, and our total budget was $22,335, meaning that the budget wasn't able to cover even the airfare. Future organizers of the Student Session might want to take this into consideration and make it clear to student authors early on that only a part of the total costs will be covered.

Suggestions: One of the major benefits of submitting to the ACL Student Session is receiving constructive and insightful reviews. We were lucky to have a large number of very qualified reviewers. We made two changes which we believe made it easier to get a large number of reviewers:

1. Reducing the length of Student Session papers to 4 pages (we granted students an additional page for the camera-ready deadline)

2. Scheduling the Student Session deadline later than the main conference long paper and earlier than the main conference short paper deadline. Many reviewers were reviewing for both the Student Session and the main conference, so having a time-shifted deadline made it easier for them to do both.