| paper | d4-conf-ijcai-2020-0397 |
| 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 gives embedding and seed-selection procedures, but no numbered theorem, lemma, corollary, or proposition asserting an algorithmic guarantee or complexity result. With no eligible computational result to continuize, the mirror grade is red.
fails bit a — no named computational result to mirror
No eligible result is covered; the paper reports methods and empirical comparisons without a numbered computational theorem.
I can’t make an anchor-based case from this paper: the text contains no numbered theorem, lemma, corollary, or proposition asserting a computational result. Its contributions are an embedding and seed-selection method, evaluated experimentally.
The introduction cites Kempe et al.’s NP-hardness result for classical influence maximization, but that result is external and unnumbered here, so it is not an eligible anchor. There are therefore no per-anchor continuous questions to state under your criteria.
The proponent raises no anchors, and I don’t see a named result to contest. The paper’s own contributions are an embedding and seed-selection method, evaluated experimentally. It states no theorem or complexity result for either. The NP-hardness claim for classical influence maximization is attributed to Kempe et al.; it is background about a different problem, not a result of this paper. On the requested anchor standard, that is decisive: there is no result here whose continuous complexity landscape this programme could chart.
The strongest possible rescue would be to study fair influence maximization on a continuum network: represent agents by latent types, model connections and cascade probabilities between types, and choose a seed mass that maximizes expected reach subject to equal per-capita influence across groups. That could be a sensible computational problem, especially for large populations with a graphon or block-model description. But it would be a new model, not a mirror of a computational result in this paper. For a general graph, a type would need to retain enough neighborhood and edge-probability information to predict diffusion; grouping nodes only by sensitive attribute or embedding can discard information that changes who is reached. Keeping each node’s full relational profile avoids that loss but usually leaves essentially one type per node. A graphon or block-model assumption makes a continuum treatment possible by imposing additional structure, and the paper gives no result about that regime.
So the negative case is strong on the paper-specific question: it offers no eligible computational result to mirror, and a continuous fair-influence problem would be a follow-up research direction rather than a continuization of one of its findings. The broader claim that *no scenario* could yield a worthwhile continuous mirror is weaker. A well-chosen graphon regime may support one; this paper simply does not establish or analyze it.
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.