Introduction

The purpose of this document is to provide sufficient guidance to ensure a successful conference.

Highlighted passages indicate items that are meant to be enforced by the General Chair. Click buttons for suggestions and opinions based on prior experience.

Part I:
Before The Papers Arrive

Conference Objectives

One of your first tasks is to determine what if any particular objectives you have for the conference. For instance, you may decide to strive to increase attendance or to bring in some related field or fields. These objectives must be reflected in the construction of your program committee and in your publicity for the conference. Your conference objectives must be communicated to and approved by the Steering Committee before commencement of the immediately prior VL Conference. Generally, you will be expected to present these objectives at the Program Committee meeting held in conjunction with the prior VL Conference. This will provide you with an opportunity to solicit suggestions for possible program committee members who can help support your objectives.

Program Committee

The program committee must be selected carefully to reinforce the general goals of the conference as well as any particular objectives of your conference. Below are some suggestions on selecting your program committee. The Program Committee must be submitted to the Steering Committee for approval by no later than three weeks following the immediately prior VL conference. Ph.D. students are not acceptable as Program Committee members.

Publicity

For publicity, there are at least three things to think about: timing, coverage versus cost, and sincerity.

Timing

What you are trying to do is get an adequate number of papers submitted within the areas you seek to promote. To do this requires early and persistent publicity. Such publicity is a continuous effort, rather than a one time shot. The following deadline for publicity is a MINIMUM guideline: (1) as soon after you have a Program Committee as possible, but no later than four months prior to the submission deadline, (2) again at approximately three months in advance incorporating an update on special events, keynote speakers, etc., and (3) again four to six weeks prior to the submission deadline. Additional notices can only help. Attached are samples of each first call, second call, and third call.

Coverage versus Cost

Your objective is to spread publicity widely, as widely as possible while staying within a budget.

It is possible to rely too much on electronic means for promotion activities. Electronic media are at times not as visible to the potential contributors/participants as the 'traditional' printed forms. Prestigious conferences continue to deliver promotion brochures, leaflets and posters.

Sincerity

Once you decide on the areas you want to attract, your call for papers must support these areas. In large part this is accomplished by having Program Committee members who are well known in each area. For new subareas, you will need to add a special note attached to the call for papers addressed particularly to the new subareas. Also the call will need to be sent especially to their mailing lists and newsgroups, telling them that yes, you mean particularly them. Without such special effort to include these new subareas, authors will not believe we want them if they haven't seen much of their area in VL recently.

The Web Page

You will need a collection of web pages to promote your conference. These pages must be up-to-date at all times, generally preceeding each public notice as well as each step in the review process. For instance, you web pages announcing the call for papers should be in place before the first call goes out.

Paper Deadlines

The deadline for papers should be no later than five months before the upcoming conference. In addition, it is recommended that when possible the deadline should precede not follow potentially competing conferences whose deadlines are close to the same time.

Keynote Speakers and Special Events

Early selection of keynote speakers is another means of reinforcing a conference theme. When possible, keynote speakers should be chosen such that their names can be announced by at least the third call. If they are known at the time of the first or second call that is even better. Just as the Program Committee selection will have an impact on paper submissions, so will the keynote speakers.

LIkewise, any special events and tutorials you plan should be arranged so that they can be announced by at least the third call for papers. These also increase interest in a conference and encourage authors to want to attend, and thus to submit a paper.

Number of Submissions

Quality is related to quantity. You need a large number of submissions if you are to pick a quality batch of papers. In general, you should target for at least three papers for every one you intend to accept. To get lots of submissions, lots of publicity is needed early. This means the biggest publicity push happens very early -- at least four months ahead of the papers deadline!

How to Attract Better Papers

Authors respond to the level of expected quality. Hence, let them know the review criteria. Publish the evaluation criteria via the web pages and in at least the second and subsequent calls. Your description does not have to be lengthy, just tell them succinctly what you're looking for.

You must also be sure to include those the same evaluation criteria in writing on the review form to be used by the Program Committee. This reminds the Program Committee members to keep in mind the things the conference publicity has been saying.

Part II:
After the Papers Come In

The Review Process

All papers to be reviewed must conform with submission length requirements, including as they affect length any specification of number of pages, treatment of figures and tables, margins, font sizes, and line spacing. Should any paper fail to comply, such paper is to be returned unreviewed, but with correspondence to the author(s) allowing N days to resubmit the paper in a revised form which conforms with the submission length requirements. The value of N is left to the discretion of each year's Program Chair(s).

All papers must have at least three reviews, and where two or more reviews differ in a substantial way, there must be a resolution of the differing views involving at least the original reviewers. In the process of resolving differing views, it may be necessary to involve yet more reviewers, but it is generally better if the original reviewers can resolve their differences. One way to handle this is to distribute to the paper's reviewers a copy of each other's comments and ask for a resolution. Generally, this will result in a discussion leading to a revision of the reviews which successfully eliminates the conflict.

The Review Form

Be specific about what you want the reviewers to comment upon. Be sure that any evaluation criteria you included on the call for papers and other publicity are reflected in the review form. Otherwise, reviewers may have quite a different idea about what is acceptable than you do. Attached is a review form sample.

The Selection Process

The selection metric you intend to use must be made known to all the Program Committee before they make their reviews. Afterwards, scores and rankings as well as tentative selections must be made available to the Program Committee before final selections are made. Comments which will improve the quality of the selection should be encouraged before final selection. This can be a tricky process as you may get comments from reviewers supporting their own papers. The key point is to be consistent and fair in your final selections.

Quality Issues in Determining Acceptance

Comparable quality conferences have acceptance rates between 20 and 30%. Keeping the quality of papers high helps to promote the conference long-term.

Accept as many posters as you can fit. The quality of poster papers does not seem to affect people's perception of quality as long as you do not allow poster presentations to exceed about 2 minutes. Some years, marginal papers have been accepted in poster form, and this seems to work out fine. Poster papers are a possible solution for marginal papers you would like to include, but do not really measure up to the quality you desire.

Demos are another means of accepting marginal papers without diluting the quality of the papers accepted.

Scheduling Issues

You must have to have a reasonably good idea of how you will schedule the conference sessions before you commit to accepting papers.

Quality of the Conference Presentation Plan

Remember to look out for the attendees, not the speakers/presenters.

The Advance Program

The Advance Program is what causes people who have been considering the conference to actually attend. It must bring out the best the conference has to offer! It highlights the special events and keynoters, waxes poetic about any social events planned, and lists the papers, including their authors and affiliations. The Advance Program must be out not less than two months after the paper deadline and at least three months prior to the conference.

Avoiding Author No-Shows

There have been years in which some authors wanted publication of their paper, but did not want to attend the conference. You must take steps to prevent this. In particular, you must require that at least one author of every accepted paper register for the conference prior to the press date of the proceedings. IEEE will help you with this, but you will have to monitor the process to make sure it is enforced.

Demos

Demos have been available in some form at almost every recent VL, and are very popular with attendees and with presenters. However, demos can be a lot of trouble if you provide the equipment, so you need to decide in advance what, if any, equipment you can supply. Also, even if there is no official demos program, it is well-received if you can set aside a room for people with laptops to use for informal demos. Also remember that demos require a lot of room, should be near the conference hall yet not so near as to disturb the speakers, and need some system managers who can help should anything go wrong with the parts you provide (eg: equipment, Internet connections, software, etc.).

Tutorials

Tutorials need to make money for the conference, so you cannot afford to hold them if there isn't a reasonable sign-up (i.e., at least enough to break even). VL is generally small as conferences go and generally cannot support more than two tutorials.

It is the customary to offer an "Intro to VLs" tutorial as one of the two. Such tutorials have been offered at VLs in the past by Ephraim Glinert, Allen Ambler, David McIntyre, Rebecca Walpole, and Margaret Burnett, and have usually been well-attended. Such a tutorial allows those new to VLs to come up to speed to some extent before the conference begins.

When you accept a tutorial, get a contract in writing with the speaker right away, with the money spelled out and a commitment. (IEEE can provide a sample contract.) Otherwise, there is the danger that misunderstandings and unexpected cancellations will occur. Publicize the tutorials widely.

Tutorials must be finalized in time that announcments can be included in the Advance Program, encouraging registrations. People need to know about them in order to make their travel plans. If the conference is held near a major population area, you can do some publicity pushing just the tutorials for those wishing to spend only a half-day on VLs, and that may raise extra money for the conference.