| paper | d4-conf-atal-2008-0317 |
| 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 has no numbered theorem, lemma, corollary, or proposition asserting a computational result. Its numbered items are definitions, while the query-answering algorithm and mapping scores appear as an algorithm and empirical evaluation, not a qualifying named result. The proposed support-share calculation is a coherent new extension, but it cannot supply the missing anchor.
fails bit a — no named computational result to mirror
No result is anchored; the paper's argumentation definitions, mapping evaluation, and query-answering procedure are not stated as numbered computational results.
The strongest honest positive case is limited by the paper’s results: it contains no numbered Theorem, Lemma, Corollary, or Proposition asserting a computational complexity result. Its numbered items are definitions; Table 4 gives a query answering algorithm without a runtime or complexity classification, and Table 3 reports empirical mapping scores. The paper says that general query answering is undecidable and describes restrictions from reference [9], but presents no numbered result establishing that claim. So there is no qualifying anchor, and I can’t honestly give an anchor-backed continuous problem or classify one by the programme’s trichotomy.
There is still a plausible *new question* suggested by the paper’s agent-consensus model. Imagine a large open e-commerce system where many agents communicate using the same pair of ontologies, but come from a small number \(\tau\) of software and policy types. A type would include its mapping approach, audience preferences, candidate arguments, and rational confidence values. Let \(\mu_t\) be the fraction of agents of type \(t\). Given candidate DL-safe query correspondences \(q\) and each type’s preferred mapping set \(P_t\), ask for the exact support share
\[
a_\mu(q)=\sum_{t:q\in P_t}\mu_t
\]
for every \(q\), and output the correspondences with positive support. These outputs are query correspondences between ontologies, as in the paper; the population is the distribution over agents. If the \(P_t\) sets are supplied, computing the shares is straightforward weighted aggregation, so this version is tractable in \(\tau\) and the number of candidate correspondences. Computing the preferred sets as part of the problem would raise the substantive complexity question, but the paper gives no named result to anchor that investigation.
The scenario is plausible in a high-multiplicity regime: many deployed agents may share the same ontologies, mapping software, and audience preferences, with \(n\) large and \(\tau\) small. It also connects to the paper’s global subjective acceptability: a correspondence has positive support exactly when some represented type accepts it. But the mass weighting adds little at that zero threshold, while requiring a positive threshold would introduce a new aggregation rule that the paper does not study.
That is the soft spot: the paper’s example has a few specialized mapping agents, not a large population, and its consensus semantics are based on whether an argument appears in agents’ preferred sets, not on how much of the population supports it. The proposed extension is therefore a credible direction for a new computational question, but not a mirror of a named computational result in this paper.
The proponent’s proposed question has no result in this paper to mirror. The paper defines its argumentation framework, reports mapping scores, and gives a query-answering algorithm, but states no numbered complexity theorem for that algorithm or for computing preferred mapping sets. Its claims about undecidability and decidability rely on prior work. That leaves room for a new question, but not an anchor-backed continuous mirror of a computational result here.
The proposed support share \(a_\mu(q)\) also adds no population-sensitive computation under the paper’s semantics. Global subjective acceptability means that some agent accepts a correspondence. With a finite support of positive-mass types, \(a_\mu(q)>0\) says exactly that some type accepts it. Repeated agents of the same complete type do not change the argument graph or that type’s preferred set. If those sets are supplied, the share is just weighted aggregation; if they are not, the task is to compute the sets once per type, and the masses do not affect that computation.
A stronger version could ask for weighted consensus, or the expected success rate of communication between randomly paired agents. Those are plausible questions, but they require a new aggregation rule or utility objective. The paper’s mappings are computed for particular ontologies and accepted through argumentation; it does not define how population shares should trade off one audience’s mapping against another’s, or how interactions should be weighted. Once such a rule is supplied, the mass calculation follows the per-type or per-pair mapping computations. Making masses drive a selection or optimization would therefore introduce a new control problem, rather than continue a computational question the paper studies.
I would not claim that no large population of similar agents is plausible. The weaker point is more consequential here: multiplicity can describe the setting, but the paper’s stated task does not use multiplicity. I can see a worthwhile research direction in weighted argumentation over agent populations, but the proponent’s example does not establish a worthwhile continuous mirror of this paper.
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.