NB: unless specified otherwise, "ICDM organizers" in this document mean the Program Chairs and Conference Chair of each year's ICDM conference.
ICDM is held every year, in different regions of the world, generally in November.
The conference program of each year consists of high-quality contributed papers on all aspects of data mining. Original (previously unpublished) manuscripts on various data mining related topics are solicited.
Location decisions are made by the ICDM Steering Committee. The Steering Committee selects the Program Chair(s).
The Steering Committee encourages proposals for location by teams vying to hold the ICDM conference. Proposals for future years' ICDM conference will be discussed at each ICDM conference.
Each proposal should have a mention of the planned social items, such as lunches, receptions, e-mail access, and possible sightseeing activities.
The organizers of each ICDM conference agree with the Steering Committee's terms on conference publications and ICDM Funds (if any).
ICDM Program Chairs of each year are advised not to submit any papers to ICDM in that particular year.
ICDM Calls for Papers.
ICDM best paper awards will be conferred on the author(s) of the Best Paper and the Best Application Paper at each ICDM conference. The awards will be announced in the calls for papers.
A selected number of ICDM accepted papers will be expanded and revised for possible inclusion in the KAIS journal (Knowledge and Information Systems, by Springer-Verlag) each year. This will be mentioned in all calls for papers of the ICDM conference.
The membership of the Steering Committee will be published in the calls for papers of ICDM conferences.
The Steering Committee should be consulted before the call for papers of a forthcoming ICDM conference is announced.
ICDM Established Researchers in Data Mining and Program Committee Formation.
For the long-term benefits of the ICDM conference organization, the ICDM Steering Committee compiles a database of "Established Researchers in Data Mining" at a password-protected website. This database is used by the ICDM organizers each year to invite suitable Program Committee members. The database is accessible to ICDM Steering Committee members only, but is also accessible to each year's ICDM organizers for that particular year.
An "Established Researcher in Data Mining" is a data mining researcher in at least one of the following 3 categories.
Serves/has served on reputed data mining related journals, such as
Has served more than once on relevant conference program committees, such as
Has published more than once in the above journals and conferences as a principal author.
The database also contains (1) information about who served on which year(s)' ICDM Program Committee, and (2) comments from each year's Program Committee Chairs for some of these established researchers if they
The Program Committee of each year must replace at least one third of the previous year's Program Committee members.
While PhD students and postdoctoral research associates are encouraged to help with ICDM paper reviews (generally with their advisors/supervisors), they are not included in our database of established researchers, and therefore are not good candidates for ICDM Program Committees.
Each year, the ICDM organizers are invited to forward their nominations for additions to the database, but before these nominations are approved (normally within 3 days) by the Steering Committee, they are not eligible for membership of the Program Committee.
Each nomination should include
The candidate's name
The candidate's e-mail address
The candidate's home page on the Web if applicable
Reasons the candidate is nominated
ICDM Paper Reviews and Selection.
ICDM encourages electronic submissions each year. The paper list (preferably with abstracts) will be put at a password-protected website after the submission deadline, so that each Program Committee member has a chance to choose their own preferences to review.
In the review form, the Program Committee members/reviewers are asked to answer
whether the paper is recommended for the best paper award, and
whether an extended version of the paper is recommended for publication in the KAIS journal.
ICDM encourages both applied and theoretical research in data mining, and promotes high quality, novel and daring research findings in every data mining area. When selecting ICDM papers, each year's ICDM organizers should consider the balance of different research styles, such as theory, methodology, algorithms, systems, integration, and applications.
Paper acceptance decisions (such as acceptance rate, and the ratio between Regular Papers and Short Papers) should be discussed with the Steering Committee.
There are reasons for not accepting papers as posters (but short papers instead), and each year's ICDM organizers are encouraged to organize a poster/demo session for every accepted paper to participate.
Papers exploring new directions will receive a careful and supportive review at ICDM conferences.
Vice Chairs are a very useful mechanism in paper reviewing and selection. The ICDM organizers of each year are encouraged to have a small number of Vice Chairs, and convene a meeting with these Vice Chairs for the final paper selection.
Decisions on invited speakers should be discussed with the Steering Committee.
There are parallel sessions at ICDM conferences, at least 2, but no more than 3.
In addition to presentations of accepted papers and invited talks, the organizers of each ICDM conference are also encouraged to organize workshops, tutorials, and data mining software/product demonstrations.
Each year's ICDM organizers are encouraged to call for or invite papers from distinguished researchers for a special session on Emerging Data Mining Technology.
The proceedings of each workshop should be published in a uniform format (using the IEEE Computer Society layout), with support from the main conference of each year.