icde-submission

ICDE Submission

Use this for an ICDE research-track submission audit. Reopen the current edition's Call for

Research Papers, Important Dates, and Submission Guidelines before giving deadline-ready

advice; ICDE re-declares its rules on a fresh site each year.

Submission audit

• Confirm the target track. ICDE runs a research track, an **industry and applications

track**, plus demos/tutorials — route with icde-topic-selection before formatting.

• Pick the round consciously. Each ICDE edition takes research papers in two rounds

(ICDE 2027: Round 1 paper deadline June 11, 2026; Round 2 November 11, 2026). A paper

rejected in Round 1 is not eligible for Round 2 — it waits for the next edition. Choose

the round you can actually be ready for, not the nearest one.

• Set up Microsoft CMT early: every co-author needs a CMT profile, and conflicts (domain

and individual) are author-declared in CMT. A missing conflict is a review-integrity issue,

not a formatting nit.

• Apply the IEEE double-column template without margin, font, or spacing edits. Long

research papers run to 12 pages plus unlimited references (short papers 6 pages plus

references, per the ICDE 2026 wording — re-confirm for the current edition).

• ICDE is single-blind: author names stay on the paper. Do not anonymize author

identity, self-citations, or the repository — that is a SIGMOD reflex, wrong here.

• Attach supplemental material (code, data, artifacts). Availability is explicitly

considered in the evaluation, so an empty supplement is a scored weakness.

• Check the original-work rule: no prior publication longer than roughly four double-column

pages covering the same work.

Blocking risks

• Submitting to a round the paper cannot be ready for, then being locked out of the same

edition's second round after a Round 1 reject.

• Overlength main text or an altered IEEE template.
• Missing or wrong CMT conflict declarations for any co-author.
• No supplemental material where the call expects availability evidence.
• A systems claim with no operating point, tuned baselines missing, or cost undisclosed.

Desk-reject and triage table

| Trigger | Severity at ICDE | Repair window |

|---|---|---|

| Overlength main text | Desk reject | None after the deadline |

| IEEE template tampering | Desk reject or chair flag | None |

| Missing CMT conflict for a co-author | Integrity flag, possible desk reject | Only before the deadline |

| Empty or unrunnable supplement | Scored weakness | Fixable before the deadline |

| Round-1 reject re-sent to Round 2 | Administrative reject | None — wait for next edition |

| Unfalsifiable performance claim | Review-stage damage | Fixable in the writing/experiments pass |

Final-week sequence for a systems paper

1. Freeze the mechanism description and the claims map so every figure has a home in the text.
2. Regenerate each figure and table from logged runs so the PDF and supplement cannot diverge.
3. Fill the supplement: run_all.sh, run_small.sh, a README that orients a reviewer in a

minute, and the workload generators with their seeds.

1. Verify every co-author's CMT profile exists and all conflicts are filed.
2. Confirm the abstract in CMT matches the PDF abstract, and the paper is within 12 pages.

Format anchors

• IEEE two-column layout is tight; wide system diagrams, algorithm blocks, and multi-series

throughput plots that fit a one-column draft routinely overflow — compress early.

• Every number in the pack (page limit, deadlines, blind policy) describes a specific edition;

treat all of them as provisional and re-check the current call.

Output format

[ICDE readiness] Ready / Needs fixes / Not ready
[Round + track] <research/industry, Round 1 or Round 2, with source date>
[Blocking checks] <round-eligibility / page / CMT-conflicts / supplement / IEEE-format>
[Evidence risk] <one systems-evaluation issue>
[Desk-reject risk] <one issue>
[Fix order] <ordered fixes before the round deadline>