Guidelines and policies

From Admin Wiki
Revision as of 17:49, 24 November 2008 by StevenBird (talk | contribs) (New page: ==Guidelines for ACL Program Committee Chairs== '''Conference Handbook - Guidelines for ACL Program Committee Chairs''' The following are guidelines for the prog...)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Guidelines for ACL Program Committee Chairs

Conference Handbook - Guidelines for ACL Program Committee Chairs

The following are guidelines for the program committee; some of it is extracted from a document prepared by Johanna Moore in 1993 summarizing issues that arose in past program committee meetings and the views of the program committees as to how they should be addressed. Other portions are the result of discussions following recent conferences.

  • Program Committee:
    • ACL has a three-tiered program committee structure: the PC Co-chairs, the area chairs, and the regular members of the PC.
    • ACL conferences work with a three-tiered program committee structure: PC Co-Chairs, Area Chairs or Senior PC Members, and regular members of the PC.
    • The Area Chairs or Senior PC members are selected by the PC Co-chairs, subject to the approval of the ACL Executive Committee. ACL policy prohibits members of the Executive Committee from serving as Area Chairs; for joint conferences with a local chapter, the same restriction applies to members of that chapter's Executive Board. Potential Area Chairs should not be contacted until their selection has been approved by the ACL Exec. In selecting Area Chairs, consideration should be given to prior experience as a regular PC member, having new people as Area Chairs, coverage of the areas of anticipated submissions, and geographical and gender distribution.
    • Area Chairs should be informed of their responsibilites (including the selection of the recipient of the Best Paper Award) at the time they are invited to serve as Area Chairs.
    • Area Chairs can cover more than one area and can overlap areas. In the Call for Papers, some past PC Co-Chairs have preferred not to list the specific areas of the Area Chairs in order to avoid biasing the submissions.
    • The Area Chairs select the regular members of the PC, in consultation with the PC Co-chairs. Consideration should be given to having a balance of experienced and new members of the PC and geotraphical and gender distribution.
    • The conference proceedings should list the PC Co-Chairs, the Area Chairs, or Senior PC Members, the Members of the Program Committee, and any additional reviewers who contributed to reviewing papers, using the headings "PC Co-Chairs", "Area Chairs" or "Senior PC Members", "Program Committee Members", and "Additional Reviewers". The names and affiliations for PC Chairs, Area Chairs or Senior PC members, and PC members should be given in the proceedings. For additional reviewers, only the name is given. All lists should be compiled in alphabetical order.
    • Each regular PC member is in charge of the review of 5-10 papers, whereas an additional reviewer reviews at most 2 papers. In the unlikely event that a reviewer is asked to review 5 or more papers he or she may, at the discretion of the Program Co-chairs, be upgraded to the status of a regular PC member. This upgrading process must be triggered by the PC member who allocated so many papers to a single reviewer, by a suitable suggestion to the Program Chair.
    • PC members may recruit additional reviewers, but the responsibility to get adequate reviews for the allocated papers remains with the PC members. The PC members report to the Area Chairs or Senior PC Members, and they in turn report to the PC Co-chairs. Additional reviewers report to the PC member who recruited them.
    • Only the PC Co-Chairs and Area Chairs or Senior PC Members can be reimbursed for attending the PC meeting. If individual PC members want to attend the PC meeting, they may do so, but on their own expense. Reviewers cannot attend the PC meeting.
  • Use of the START conference software:
    • For the past several years, ACL has used the START conference software for paper submission, reviewing, and helping with the publication process. The PC Co-chairs should consult with the General Chair and the Executive Committee about whether a contract with START currently exists or whether a new contract should be negotiated. This needs to be settled very early, so that the conference software will be available for submissions.
  • Relevance of papers to ACL:
    • In recent years, a broader range of papers have been solicited for ACL conferences, including papers in information retrieval, spoken language recognition and understanding, tools and language resources, and applications. ACL wants to continue to encourage papers in all areas related to language processing.
    • The authors have the burden of explaining a paper's relevance to the CL community.
  • Mentoring service: ACL conferences provide a mentoring service to help authors, from countries where English is not the language of scientific exchange, produce a fluent paper. The PC Co-Chairs are responsible for setting up the mentoring service for the conference.
    • The mentoring service typically is chaired by an individual who is not an Area Chair. The Chair of the mentoring service identifies a set of mentors. Papers are sent to the Chair of the mentoring service some weeks prior to the conference submission deadline, and the mentors work with the authors to improve the communication of their research.
    • The mentoring service should be announced in the Call for Papers and publicized on the conference web site --- see sample mentoring announcement
  • Paper submission deadline:
    • Every year, the PC chair receives several papers that arrive after the deadline. Although there have been variations in how such papers are handled, the majority opinion is that the deadline should be strictly enforced in the interest of fairness to those authors who could have improved their papers if they had extra time.
    • The policy on late submissions should be explicitly stated and well-publicized. It is important that the precise time zone for the deadline be clearly stated, such as 5pm Eastern Standard Time.
    • With the electronic submission of papers, it is now easy to enforce deadlines by terminating the submission process.
  • Double blind reviewing:
    • ACL requires that papers be blind reviewed. Consequently, clear instructions must be provided about omitting references that will disclose the author's identify.
    • The PC chairs must decide how to handle papers that violate author anonymity for blind reviewing. In recent years, such papers have been returned to the authors without review.
  • Length restrictions on papers:
    • Length restrictions: Length restrictions should be stated in terms of font size, margin size, page size, and number of pages. The provision of style files makes it easy for the authors to adhere to these guidelines. Sample style files from past conferences are accessible form the Program Chair section of the Conference Handbook.
    • The PC chairs must decide how to handle papers that violate the length restrictions on submissions. It is generally felt that length restrictions must be adhered to, and that papers that violate these restrictions should not be reviewed; the basis for this view is that most authors would have liked extra pages to explain their work, but have worked hard to cut their papers down to the requisite length. Thus reviewing papers that violate the length restrictions gives an advantage to the authors of those papers.
  • Double submission policy:
    • Papers presented at an ACL conference must represent new work that has not been previously published. It is the responsiblity of the author to inform the program chair of any potential problem with respect to this requirement, as noted in the following guidelines.
    • Papers that have appeared at a conference with published proceedings constitute previously published work.
    • Papers that overlap other papers that have appeared at a conference with published proceedings must contain significant new results. Authors must include on the title page a list of any previous papers that the current paper overlaps or extends, and must identify the significant new results contained in the new submission. The program co-chairs have the final decision about what constitutes significant new results.
    • Papers that have appeared at a workshop do not constitute previously published work, as long as the paper submitted to ACL is an extension of the workshop paper. Extensions might include new results, more in-depth analysis, evaluation that was not part of the workshop paper, or further experiments. Authors must include on the title page a list of any previous workshop papers that the current paper extends, and must identify how the current submission extends the previous workshop papers. The program co-chairs have the final decision about whether the ACL submission represents an extension of the workshop papers.
    • Papers being submitted both to ACL and another conference or workshop must note this on the title page. (This includes submissions that are extensions of papers currently being submitted to a workshop.) A paper to be presented at ACL must be withdrawn from other conferences and workshops.
  • Implementation of ACL double submission policy: Papers being submitted both to ACL and another conference or workshop must:
    • Note on the title page the other conference or workshop to which they are being submitted. (This includes submissions that are extensions of papers currently being submitted to a workshop.)
    • State on the title page that if the paper is accepted for ACL, then the paper will be withdrawn from other conferences and workshops.
  • Enforcing the ACL double submission policy: Authors are expected to adhere to the ACL policy on double submission of papers. However, double submissions that do not following this policy have been problematic at recent conferences. For ACL-05, the program chairs consulted with the program chairs of other major conferences (such as IJCAI) to determine when authors had submitted to multiple conferences. The submissions were checked, and papers that were multiply submitted (ie., not substantially different) and accepted by more than one conference were rejected from one or both of the conferences.
  • Paper reviewing and acceptance:
    • ACL Policy on Conflicts of Interest:
      • ACL has adopted a formal conflict-of-interest policy. It is essential that the PC co-chairs, area chairs, and reviewers be familiar with the policy.
        • Submissions by the PC Co-chairs or their students pose special problems. Although there is not an ACL policy prohibiting the PC Co-chairs from submitting to the ACL conference for which they are serving as PC Co-chairs, PC Co-chairs and their students are encouraged to submit papers to a different conference to avoid the appearance of a conflict of interest in the paper acceptance process.
    • Review forms from recent conferences are available: review form from ACL-04 and review form from ACL-05. Although the content of the review form is at the discretion of the PC Co-Chairs, the Executive Committee suggests the following in order to provide a good evaluation of application and tools papers:
      • "Does this paper provide data, software, or source code that will be useful to other researchers in this area?"
      • "How useful to the community are any resources released along with the paper."
    • The PC co-chairs and area chairs or senior PC members are responsible for ensuring a high quality technical program. Since the quality of submissions is not uniform from year to year, the acceptance rate for papers may vary from one ACL conference to another. As a guide, an acceptance rate below 20% would be unusually low and an acceptance rate of 30% would be unusually high; however, it is the perogative of the PC co-chairs to evaluate the quality of the submissions and, in consultation with the General Chair, to determine what is an appropriate acceptance rate for the particular conference.
  • Best Paper Award:
  • Visas: Although there has been concern about visa problems preventing researchers from attending conferences in the United States, it does not appear that there were more visa rejections for ACL-05 than there have been for ACL conferences in other countries. To address the visa problem, the PC Co-chairs should notify the authors of submitted papers that if a visa will be needed to attend the conference, authors should do the following:
  1. set up an appointment with the appropriate consulate for shortly after the acceptance notification date, and
  2. contact Priscilla Rasmussen for a letter of invitation when the paper is accepted.