Pith. sign in

REVIEW 2 major objections 5 minor 24 references

An Extension-Based Argument-Ranking Semantics: Social Rankings in Abstract Argumentation Long Version

T0 review · 2 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read By applying social ranking functions to extension rankings, an argument-ranking semantics can refine the classical skeptical/credulous/rejected acceptance classification, and the paper pins down the exact axioms on the social ranking that…

desk verdict A solid bridge between extension-based and ranking-based argumentation, with a local proof gap in the main theorems that is repairable and does not threaten the core construction. read the letter →

arxiv 2412.13632 v1 pith:TH7FSXJ3 submitted 2024-12-18 cs.AI

Lars Bengel , Giovanni Buraglio , Jan Maly , Kenneth Skiba This is my paper · ORCID
classification cs.AI
keywords abstractargumentationargument-rankingsemanticsextension-rankingsocialrankingfunctionslexicographicexcellenceacceptancerefinementaxiomaticproperties
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes a new family of argument-ranking semantics built by feeding an extension-ranking semantics into a social ranking function. Its central result is a pair of sufficient conditions: if the extension ranking generalises a classical semantics and the social ranking satisfies Independence from the worst set and Pareto-efficiency, then the resulting argument ranking places every skeptically accepted argument above every credulously accepted argument, and every credulously accepted argument above every rejected one. The paper also proves necessary conditions in terms of a Dominating set axiom and shows that the lexicographic excellence operator (lex-cel) satisfies the sufficient conditions, making it a concrete instance of the construction. The upshot is a principled way to order arguments within the acceptance classes, not just across them.

What carries the argument

The central machinery is the rank of a set with respect to a preorder over the powerset, defined as the length of a longest strict descending chain ending at that set, together with the lexicographic excellence (lex-cel) social ranking function, which compares two elements by the number of sets of each rank they belong to, moving lexicographically from the best rank downward. The transfer theorems work by constructing an auxiliary two-level preorder that separates extensions from non-extensions, using Pareto-efficiency to establish a strict preference inside that auxiliary ranking, and then invoking Independence from the worst set to carry the strict preference over to the actual extension ranking. The Dominating set axiom, implied by the two sufficient axioms, captures the intuition that an element contained in a set that dominates every set containing a rival element must be ranked above that rival.

What would settle it

Take any argumentation framework with at least one skeptically accepted argument and at least one credulously but not skeptically accepted argument, apply lex-cel with the complete extension ranking, and check whether the skeptically accepted argument is ranked strictly above the credulously accepted one; a single framework where this fails would refute the claim that this semantics satisfies skeptical-Compatibility. Alternatively, re-run the proof of Theorem 2 after adding reflexivity to the auxiliary relation and verify that the Pareto-efficiency step still yields the strict preference; if it does not, the transfer argument fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that any social ranking function can be turned into an argument-ranking semantics by applying it to an extension-ranking semantics, and that the right choice of axioms on the social ranking makes the resulting semantics a true refinement of the classical acceptance classification. In Theorems 2 and 3 it proves that if the extension-ranking semantics satisfies generalisation and the social ranking satisfies Independence from the worst set and Pareto-efficiency, then the induced semantics satisfies Compatibility and skeptical-Compatibility; hence skeptically accepted arguments are ranked above credulously accepted ones, which are ranked above rejected ones. The paper then shows that lex-cel satisfies Pareto-efficiency and, together with the complete extension ranking, yields the concrete ranking of the running example a > d > c > b. It also establishes necessary conditions: Dominating set is implied by the two sufficient axioms and is itself necessary for a social ranking to yield conflict-free Compatibility and admissible Compatibility.

Load-bearing premise

The transfer arguments in Theorems 2 and 3 require the auxiliary two-level relation separating extensions from non-extensions to be a preorder, but as written it is not reflexive, and without adding equality the Independence from the worst set step does not go through.

Editorial extensions

If this is right

  • Under the sufficient conditions, the induced semantics orders every skeptically accepted argument strictly above every credulously accepted argument, and every credulously accepted argument strictly above every rejected one.
  • The lexicographic excellence operator applied to the complete extension ranking satisfies the refinement property and is strictly more informative than the earlier extension-ranking-based argument-ranking semantics, because a lex-cel preference implies a preference in that earlier semantics.
  • The Dominating set axiom is necessary as well as sufficient on realisable preorders: any social ranking that, combined with the conflict-free or admissible extension rankings, yields the corresponding compatibility principle must satisfy Dominating set.
  • The serialisability-based ranking semantics, the only previously known semantics satisfying admissible Compatibility, violates complete skeptical-Compatibility, so the lex-cel construction is the only known argument-ranking semantics satisfying the refinement property.
  • The singleton approach, which ranks arguments by comparing only their singleton sets, does not refine the acceptance classification and therefore cannot deliver the refinement property.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same two-level auxiliary-preorder transfer technique should generalise beyond argumentation: any domain where the top level of a preorder is the set of 'winning' coalitions can use Independence from the worst set to lift a Pareto-derived strict preference over individuals to the full ranking.
  • The written proof of the transfer theorems uses a non-reflexive relation as a preorder; adding equality repairs the gap, so the theorems survive but the appendix should be read with that patch in mind.
  • The necessity results suggest a design principle for future social-ranking-based semantics: check the Dominating set axiom first, since it is exactly what buys the coarse acceptance refinement, and then add axioms to shape the fine-grained ordering within the acceptance classes.
  • A testable extension would be to apply other social ranking functions satisfying the two sufficient axioms to the same extension rankings and compare the fine-grained orders they induce within the credulous and skeptical classes on benchmark frameworks.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper proposes a new family of argument-ranking semantics for abstract argumentation, obtained by applying social ranking functions to extension-ranking semantics. It introduces a rank-based generalization of social ranking functions to partial orders, defines the lexicographic excellence operator (lex-cel), and studies when the induced argument-ranking semantics refines the classical skeptical/credulous/rejected classification. The main results (Theorems 2 and 3) show that if the extension-ranking semantics satisfies sigma-generalisation and the social ranking function satisfies Independence from the Worst Set and Pareto-efficiency, then the induced semantics satisfies sigma-Compatibility and sigma-skeptical-Compatibility, hence sigma-Refinement. The paper also provides necessary conditions for related compatibility properties in terms of a Dominating Set axiom, and shows that lex-cel satisfies Pareto-efficiency, making lex-cel_tau a concrete instance satisfying sigma-Refinement. The proofs are given in an appendix; the paper reports no experiments and no fitted parameters.

Significance. If the central results are correct, the paper offers a principled and modular bridge between extension-based argumentation semantics and ranking-based semantics, with a clear axiomatic decomposition: the refinement property is traced to Independence from the Worst Set and Pareto-efficiency of the social ranking function, rather than to ad-hoc constructions. The use of existing published definitions, the absence of free parameters, and the explicit statement of proofs are strengths. The paper's main claims are, however, presented more strongly than what is proven: the body gives sufficient conditions for the refinement property and necessary conditions for a closely related property, not a necessary-and-sufficient characterization, and the proofs of the central theorems contain a technical gap concerning the use of a non-preorder auxiliary relation. These issues are repairable, but they currently affect the validity of the paper as written.

major comments (2)
  1. [Section 4, Theorems 2 and 3] The proofs of Theorems 2 and 3 introduce the auxiliary relation ⊒σ defined by X ⊒σF Y iff X ∈ σ(F) and Y ∉ σ(F). This relation is not reflexive and therefore not a preorder; however, Definition 10 (rank), Definition 12 (Independence from the Worst Set), and Definition 13 (Pareto-efficiency) are all stated only for preorders on P(S). Consequently, the applications of rank⊒σ, IWS, and Pareto-efficiency to ⊒σ are not legitimate as written. Additionally, the proof of Theorem 3 asserts that τ can be viewed as a refinement of τ′, which is false in general: σ-generalisation only identifies the maximal elements of τ with σ(F), and a non-maximal set can be incomparable to a maximal set in a preorder. These gaps are repairable by taking the reflexive closure of ⊒σ, which is a two-level total preorder with top level exactly σ(F), and by verifying that the rank-1 sets and strict preferences are preserved so that IWS applies; with this repair the theorems' conclusions follow. As submitted, however, the proof of the central result is technically incomplete.
  2. [Abstract and Section 1] The abstract claims that the paper provides 'necessary and sufficient conditions for a social ranking function to give rise to an argument-ranking semantics satisfying the desired refinement property,' and Section 1 repeats that the axiomatic properties shown are 'sufficient and necessary.' The body, however, proves sufficient conditions for σ-Compatibility and σ-skeptical-Compatibility (Theorems 2 and 3) and separate necessary conditions for the weaker property cf-C/ad-C in terms of the Dominating Set axiom (Theorems 5 and 6). No theorem in Section 4 establishes a single condition (or conjunction) that is both necessary and sufficient for the full σ-Refinement property. The claim should be weakened to 'sufficient conditions' plus 'necessary conditions for related principles,' or a genuine characterization of σ-Refinement must be supplied.
minor comments (5)
  1. [Section 4, Theorem 5 proof] In the proof of Theorem 5, the sentence 'As X contains x, its set of conflicts must be a strict super-set of the conflicts in {x}' is false; a superset X of {x} can have the same conflict set as {x}, e.g., when both are conflict-free. The desired conclusion {x} ⊉ Y actually follows from CFF({x}) ⊆ CFF(X) ⊂ CFF(Y), which is obtained from the assumption X ⊒ Y, so the proof is fixable but the written argument is inaccurate.
  2. [Appendix, Proposition 4 proof] In the proof of Proposition 4, the displayed set expression '{Z ∈ P | x, y ⁄∈Z ∧ rankr-cf(Z ∪ {a})} < rankr-cf(Z ∪ {b}) = {∅}' is malformed; it should read {Z ∈ P | x, y ∉ Z ∧ rankr-cf(Z ∪ {a}) < rankr-cf(Z ∪ {b})} = {∅}.
  3. [Example 3] The phrase 'the final admissible sets ∅ and {d}' is unclear; 'final' should be replaced by 'only remaining' or a similar expression, since these are the admissible sets that are not complete extensions.
  4. [Definition 10] The definition of rank relies on the existence of a longest strict chain, which is guaranteed only if the underlying set S is finite (or the preorder has no infinite descending chains); the paper should state explicitly that S is assumed finite, which is consistent with the finite AFs considered later.
  5. [Section 5] The claim that lex-cel_tau is 'the only known argument-ranking semantics' satisfying σ-C and σ-sk-C is a statement about the literature and should be phrased more cautiously (e.g., 'to the best of our knowledge') or supported by a more systematic comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central theorems derive the refinement property from stated axiomatic assumptions and prior, externally published definitions.

full rationale

The derivation chain is self-contained relative to its stated assumptions. Theorems 2 and 3 assume an extension-ranking semantics τ satisfying σ-generalisation and a social ranking function ξ satisfying Independence from the worst set and Pareto-efficiency, and prove σ-Compatibility and σ-skeptical-Compatibility for ξτ. The key auxiliary relation ⊒σ is defined by X ⊒σ_F Y iff X∈σ(F) and Y∉σ(F), and the proof transfers strict preferences from this two-level relation to τ via Independence from the worst set; this is an application of the axioms, not an assumption of the conclusion. The use of Skiba et al. (2021) extension-ranking semantics is legitimate prior work, not a self-citation load-bearing premise, and lex-cel is imported from Bernardi et al. (2019) rather than being fitted to the target result. No parameter is fitted to data and no known result is renamed. The only substantive technical issue is that the auxiliary relation ⊒σ, as written, is not reflexive and therefore not a preorder, so Definition 10 and the IWS/Pareto-efficiency axioms cannot be applied to it literally; this is a local proof gap repairable by taking the reflexive closure, and it does not make the derivation circular. Consequently the central claim has independent content and the circularity score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new entities or fitted parameters. It relies on standard finiteness assumptions for abstract argumentation, on sigma-generalisation of extension-ranking semantics, and on the axiomatic framework of social ranking functions. The main unstated technical assumption is that the auxiliary two-level relation in Theorem 2 and 3 is a preorder, which is not reflexive as written.

assumptions (4)
  • domain assumption The extension-ranking semantics tau must satisfy sigma-generalisation: the most plausible sets under tau coincide with the sigma-extensions.
    Used in Theorems 2 and 3 to identify rank-1 sets with sigma-extensions; for lex-cel on the complete extension ranking this is inherited from Skiba et al. (2021).
  • standard math Argumentation frameworks are finite, so powersets are finite and ranks via longest strict chains are well-defined.
    Definition 10 and all counting arguments assume finite rank sequences; the paper only considers finite AFs in Section 2.
  • domain assumption Social ranking functions take a preorder on the powerset as input; the paper extends this to partial orders via the rank-of-a-set notion.
    Definition 10 and Definition 16 assume the extension-ranking semantics yields a preorder on the powerset, and that ranks can be computed from longest strict chains.
  • domain assumption Pareto-efficiency and Independence from the worst set are coherent axioms for social ranking functions.
    These axioms are borrowed or introduced as premises in Theorems 1, 2 and 3; the paper argues lex-cel satisfies them.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An Extension-Based Argument-Ranking Semantics: Social Rankings in Abstract Argumentation Long Version." pith.science (2026). https://pith.science/paper/TH7FSXJ3

@misc{pith2026241213632,
  author       = {Pith},
  title        = {Pith review of: An Extension-Based Argument-Ranking Semantics: Social Rankings in Abstract Argumentation Long Version},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TH7FSXJ3}},
  note         = {Machine review of arXiv:2412.13632}
}
read the original abstract

In this paper, we introduce a new family of argument-ranking semantics which can be seen as a refinement of the classification of arguments into skeptically accepted, credulously accepted and rejected. To this end we use so-called social ranking functions which have been developed recently to rank individuals based on their performance in groups. We provide necessary and sufficient conditions for a social ranking function to give rise to an argument-ranking semantics satisfying the desired refinement property.

Figures

Figures reproduced from arXiv: 2412.13632 by the authors.

Figure 1
Figure 1. Abstract argumentation framework F1 from Ex￾ample 1. to attacks R = {(a, b),(b, c),(c, d),(d, c)}. We see that F1 has three complete extensions {a}, {a, c} and {a, d}, where only the last two are preferred. In addition, we see that, a ∈ skco(F1), c, d ∈ credco(F1), and b ∈ rejco(F1). An isomorphism γ between two AFs F = (A, R) and F ′ = (A′ , R′ ) is a bijective function γ : F → F ′ such that (a, b) ∈ R iff (γ(a), γ… view at source ↗
Figure 2
Figure 2. AF F2 from Example 7. A number of other argument-ranking semantics were in￾troduced in the literature (for an overview see Bonzon et al. (2016)). However, the only known argument-ranking seman￾tics satisfying ad-Compatibility is the serialisability-based argument-ranking semantics (ser) by Bl ¨umel and Thimm (2022). The serialisability-based argument ranking seman￾tics ranks arguments according to the number of conf… view at source ↗
Figure 3
Figure 3. AF F3 from Example 8. x ≻lex-cel ⊒ y it remains to show that xi,⊒ = yi,⊒ for all i < k. By construction, for all i < k and Z ∈ P \ {x, y}, we know that rank⊒(Z ∪ {x}) = rank⊒(Z ∪ {y}). Hence, for each set containing x there is exactly one set containing y. By Definition 10, we obtain xi,⊒ = yi,⊒, as desired. 5 Related Work In the following, let A be an arbitrary set of objects and ⊒ is a preorder on the powerset P(A… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 22 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Algaba, E.; Moretti, S.; R \' e mila, E.; and Solal, P. 2021. Lexicographic solutions for coalitional rankings. Social Choice and Welfare, 57(4): 817--849

    work page 2021

  4. [4]

    Amgoud, L.; and Ben - Naim, J. 2013. Ranking-Based Semantics for Argumentation Frameworks. In Scalable Uncertainty Management - 7th International Conference, SUM 2013 , 134--147. Springer

    work page 2013

  5. [5]

    Amgoud, L.; Ben - Naim, J.; Doder, D.; and Vesic, S. 2016. Ranking Arguments With Compensation-Based Semantics. In Principles of Knowledge Representation and Reasoning: Proceedings of the Fifteenth International Conference, KR 2016 , 12--21. AAAI Press

    work page 2016

  6. [6]

    Amgoud, L.; and Beuselinck, V. 2023. An Equivalence Class of Gradual Semantics. In Symbolic and Quantitative Approaches to Reasoning with Uncertainty - 17th European Conference, ECSQARU 2023 , 95--108. Springer

    work page 2023

  7. [7]

    Baroni, P.; Caminada, M.; and Giacomin, M. 2018. Abstract Argumentation Frameworks and Their Semantics. In Handbook of Formal Argumentation, 157--234

    work page 2018

  8. [8]

    Bernardi, G.; Lucchetti, R.; and Moretti, S. 2019. Ranking objects from a preference relation over their subsets. Social Choice and Welfare, 52(4): 589--606

    work page 2019

Show all 24 references
  1. [9]

    Bl \" u mel, L.; and Thimm, M. 2022. A Ranking Semantics for Abstract Argumentation Based on Serialisability. In Computational Models of Argument - Proceedings of COMMA 2022 , 104--115. IOS Press

    2022

  2. [10]

    Bonzon, E.; Delobelle, J.; Konieczny, S.; and Maudet, N. 2016. A Comparative Study of Ranking-Based Semantics for Abstract Argumentation. In Proceedings of the Thirtieth AAAI Conference on Artificial Intelligence 2016 , 914--920. AAAI Press

    2016

  3. [11]

    Caminada, M. W. A.; Carnielli, W. A.; and Dunne, P. E. 2012. Semi-stable semantics. J. Log. Comput., 22(5): 1207--1254

    2012

  4. [12]

    Cayrol, C.; and Lagasquie - Schiex, M. 2005. Graduality in Argumentation. J. Artif. Intell. Res., 23: 245--297

    2005

  5. [13]

    Dung, P. M. 1995. O n the A cceptability of A rguments and its F undamental R ole in N onmonotonic R easoning, L ogic P rogramming and n- P erson G ames. Artificial Intelligence

    1995

  6. [14]

    Haret, A.; Khani, H.; Moretti, S.; and \" O zt \" u rk, M. 2018. Ceteris paribus majority for social ranking. In Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, IJCAI 2018 , 303--309. ijcai.org

    2018

  7. [15]

    Heyninck, J.; Raddaoui, B.; and Stra er, C. 2023. Ranking-based Argumentation Semantics Applied to Logical Argumentation. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, IJCAI 2023 , 3268--3276. ijcai.org

    2023

  8. [16]

    Khani, H.; Moretti, S.; and \" O zt \" u rk, M. 2019. An Ordinal Banzhaf Index for Social Ranking. In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, IJCAI 2019 , 378--384. ijcai.org

    2019

  9. [17]

    Maly, J.; and Wallner, J. P. 2021. Ranking Sets of Defeasible Elements in Preferential Approaches to Structured Argumentation: Postulates, Relations, and Characterizations. In Thirty-Fifth AAAI Conference on Artificial Intelligence, AAAI 2021 , 6435--6443. AAAI Press

    2021

  10. [18]

    Moretti, S.; and \" O zt \" u rk, M. 2017. Some Axiomatic and Algorithmic Perspectives on the Social Ranking Problem. In Algorithmic Decision Theory - 5th International Conference, ADT 2017 , 166--181. Springer

    2017

  11. [19]

    Moulin, H. 2004. Fair Division and Collective Welfare. MIT Press

    2004

  12. [20]

    Skiba, K. 2023. Bridging the Gap between Ranking-based Semantics and Extension-ranking Semantics. In Proceedings of the 9th Workshop on Formal and Cognitive Reasoning co-located with the 46th German Conference on Artificial Intelligence (KI 2023) , 32--43. CEUR-WS.org

    2023

  13. [21]

    Skiba, K.; Rienstra, T.; Thimm, M.; Heyninck, J.; and Kern - Isberner, G. 2021. Ranking Extensions in Abstract Argumentation. In Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, IJCAI 2021 , 2047--2053. ijcai.org

    2021

  14. [22]

    Suzuki, T.; and Horita, M. 2024. Consistent social ranking solutions. Social Choice and Welfare

    2024

  15. [23]

    van der Torre, L.; and Vesic, S. 2017. The Principle-Based Approach to Abstract Argumentation Semantics. FLAP , 4(8)

    2017

  16. [24]

    Yun, B.; Vesic, S.; Croitoru, M.; and Bisquert, P. 2018. Viewpoints Using Ranking-Based Argumentation Semantics. In Computational Models of Argument - Proceedings of COMMA 2018 , 381--392. IOS Press

    2018

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.