| paper | d4-conf-aaai-2013-0004 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
The paper states no numbered theorem, lemma, corollary, or proposition with an algorithmic or hardness claim; Algorithm 1 has no proved guarantee, and the Enron comparison is empirical. The proposed multi-type blocking model may be a sensible future problem, but it cannot supply the required computational result of this paper.
fails bit a — no named computational result to mirror
No named computational result is covered; the greedy heuristic has no proved guarantee, and the Enron comparisons are empirical.
This paper has no qualifying named anchor: it contains no numbered theorem, lemma, corollary, or proposition asserting a complexity or algorithmic result. Equation (1) defines the blocking problem, and Algorithm 1 names the greedy heuristic, but the paper proves no guarantee for it. Its evidence is experimental performance on the Enron network. So I cannot give a continuous problem *per named result* without inventing an anchor.
The strongest unanchored research lead is a multi-type continuous version of the paper’s underlying blocking problem. Let \(R\) be a fixed set of user roles, with population shares \(\mu_r\), initially infected shares \(i_r\), and transmission rates \(\beta_{rs}\) between roles. Given a blocking budget \(\kappa\), choose blocked masses \(b_r\) with \(0\le b_r\le\mu_r-i_r\) and \(\sum_r b_r\le\kappa\), to minimize the final infected mass under a multi-type Independent Cascade limit. A plausible regime is a large organization or platform with many users in a small number of roles and contact probabilities determined by role pairs. This keeps the paper’s goal—blocking uninfected users to reduce eventual contamination—and makes mass the decision unit.
The weakness is that the paper studies an arbitrary fixed graph. Role masses and transmission rates discard much of that graph’s structure, so this model requires a substantive regime assumption. It is a plausible problem for future analysis, but the paper supplies no named result that anchors it or indicates whether it is tractable.
There is no named computational result here to mirror. Equation (1) defines the blocking problem, and Algorithm 1 gives a greedy heuristic, but the paper proves no complexity bound, approximation guarantee, or optimality result for it. Its evidence is an experiment on one network. The proposed continuous problem is therefore a new research lead, not a continuization of a result established by this paper.
The strongest version of that lead would use a large block-structured network: finitely many roles, population and infected shares \(\mu_r\) and \(i_r\), and transmission rates between roles. Blocking masses \(b_r\) under a budget \(\kappa\) is a sensible, non-individualized decision problem. But to get a continuous limit from it, one must assume that users in each role mix according to those rates. That yields a multi-type epidemic-control model; it does not preserve the arbitrary graph structure that drives \(\sigma\) in this paper. Preserving that structure with a graphon-like model is another possibility, but then the “types” encode network positions and may not form a small, fixed set of interchangeable user roles. Either route needs substantial modeling assumptions that the paper neither studies nor supports with a proved guarantee for its greedy method.
So the negative case is strong against treating this paper as evidence for a result-anchored continuous mirror: no such result appears in the text. It is weak as a universal claim that no worthwhile mirror exists. A role-based population model is plausible and computationally interesting; the honest objection is that it is not anchored by a named result 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.