| paper | d4-conf-aaai-2004-0155 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
The paper’s own algorithm and running-time analysis are unnumbered, while Properties 1 and 2 state consistency relationships rather than computational results. Under the required anchor rule, no named computational result remains to support a continuous-population mirror.
fails bit a — no named computational result to mirror
No population mirror is anchored; the paper’s DTPP and TCSPP algorithm and its running-time analysis are unnumbered, and Properties 1 and 2 are structural consistency statements.
I can’t make the requested anchor-based positive case. The paper has no numbered Theorem, Lemma, Corollary, or Proposition asserting a computational complexity result. Properties 1 and 2 are numbered, but they state consistency relationships between projected STPs; they do not establish that a problem is in P or hard.
The paper does give an unnumbered analysis of its DTPP algorithm’s running time, and it cites NP-hardness of general DTP consistency from Dechter, Meiri, and Pearl (1991). Those are not qualifying in-paper anchors under your rules. So there is no valid anchor here from which to state one continuous problem per result.
The positive case names no result to mirror, so there is no anchor to defeat one by one. The paper’s numbered Properties 1 and 2 describe how consistency behaves across preference levels; they are not computational complexity results. Its NP-hardness statement concerns earlier work on DTP consistency. The paper’s own contribution is an algorithm and an unnumbered running-time analysis. That leaves no qualifying named computational result under the requested standard.
The strongest rescue is to treat that algorithm as an implicit anchor and imagine many people following temporal plans. But the algorithm solves one jointly constrained plan: its variables are named event times, its disjunctions link particular events, and its output is a schedule satisfying the whole network. For independent copies of a plan, adding population mass does not change consistency or each copy’s WLO optimum. Aggregating WLO across copies gives the worst value among positive-mass types, so the weights still do no work. A continuum would also make WLO an essential infimum, which can ignore exceptional measure-zero types. A share-sensitive population objective would require changing WLO.
A more compelling cohort-scheduling version could make mass matter through shared appointment capacity, congestion, or fairness. That is a plausible neighboring research direction. Those interactions, however, add the central coupling constraints and population objective that this paper does not study; they do not give a continuous-population version of its algorithmic result. The case against is strongest on this mismatch and the missing named anchor. It is not a proof that the authors could never recognize a cohort-scheduling model as their problem, but this paper supplies little basis for claiming that such a model would be a worthwhile mirror.
The adversarial triple: the proponent anchors on up to three named results; the opponent sees that case and must defeat every anchor; the judge decides which case convinced it. These are the pipeline’s own outputs, generated by tools/triple_run.py — no human edited them. The paper’s own text is not reproduced here beyond the quoted statement above.