Pith. sign in

REVIEW 3 major objections 4 minor 56 references

Defining Decentralization: An Ontological Perspective

T0 review · 3 major / 4 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 The paper defines decentralization as a subject-specific relational property: a subject is decentralized exactly when its projection multiplicity exceeds one, and distributed exactly when its realizations span more than one vertex.

desk verdict A genuinely useful δ/λ separation and a careful literature dissection, but the central definition rests on an unconstrained analyst-chosen realization count and the headline metric equation contradicts its own worked examples. read the letter →

arxiv 2608.09748 v1 pith:AZ5CE6EJ submitted 2026-08-10 cs.DC cs.AIcs.LGcs.LOcs.SYeess.SY

Jakub Kacper Szeląg , Aydin Abadi , Mohammad Naseri This is my paper · ORCID
classification cs.DCcs.AIcs.LGcs.LOcs.SYeess.SY
keywords decentralizationontologygraphrepresentationformalsemanticsfederatedlearningblockchaindistributionvsmetrics
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

Decentralization has been used for half a century in computer science without a transferable definition, so the same architecture can be called centralized in one paper and decentralized in another. The paper proposes to close this gap with an ontology for graph-representable communication systems in which decentralization is always decentralization of some subject—data, model training, aggregation authority, ledger replication—rather than a property of a whole system. A subject is decentralized when its projection multiplicity is greater than one, and distributed when its realizations occupy more than one vertex, with the two properties formally separated. The framework derives system-level categories (centralized, partially decentralized, fully decentralized), two graph metrics (Void Tolerance and Imperviousness), and instantiates them on federated learning and blockchain architectures to show consistent, comparable assessments.

What carries the argument

The load-bearing object is the projection particular $p_u$, an individual in the ontology that binds a system $s$, a topology $T$, and a subject $u$ through the relations hasProjection, ofSubject, and inTopology. Each projection carries a finite set of realization particulars, mapped by realizedAt to vertices of the graph; $\delta(p_u)$ counts realizations and $\lambda(p_u)$ counts the distinct supporting vertices, and the axiom $1 \leq \lambda(p_u) \leq \delta(p_u)$ keeps the two notions ordered. The logical layer axiomatizes these relations and derives the centralized/decentralized and distributed/undistributed predicates from $\delta$ and $\lambda$. The analytical layer adds two subject-specific graph metrics—Void Tolerance, which measures vertex-removal resilience of the subject-induced subgraph, and Imperviousness, which measures the edge-deletion effort needed to isolate a subject-supporting vertex—combined into a per-subject decentralization vector in $[0,1]^2$.

What would settle it

If two analysts choose equally defensible but different subject sets for the same graph, the framework returns different system-level classifications; observing such a disagreement in practice would show that the definition is not actually independent of the analyst's modeling choices.

Watch

Extended reading notes

Core claim

The central claim is that decentralization can be defined without reference to any application domain: fix a graph-representable system, choose a subject of decentralization, and let $p_u$ be the unique projection particular linking the system, its topology, and that subject. The subject is decentralized iff the number of distinct realization particulars $\delta(p_u)$ is greater than one; it is distributed iff the number of vertices supporting those realizations $\lambda(p_u)$ is greater than one. Centralization is the degenerate zero-dimensional case in which every projection has multiplicity exactly one, and a system's dimensionality is the count of its decentralized subjects. Full, partial, and centralized system classes follow from whether all, some, or none of the declared subjects are decentralized, so the same topology can legitimately yield different classifications for different subjects. The paper argues this resolves the Decentralization Problem by making the previously implicit choice of dimension explicit and formal.

Load-bearing premise

The framework's classifications depend on the analyst's choice of subjects of decentralization and on the graph abstraction; different analysts may select different subjects, and the same topology can be classified fully, partially, or non-decentralized depending on that choice.

Editorial extensions

If this is right

  • Vanilla federated learning with one aggregation server is classified as partially decentralized: data and training are decentralized across clients, but aggregation authority is centralized.
  • A blockchain with a single block-ordering node but replicated ledger and client-side transaction submission is partially decentralized; re-assigning consensus to the core nodes makes it fully decentralized under the framework.
  • A subject can be decentralized and undistributed at the same time when several realizations sit on one vertex, so decentralization does not imply distribution.
  • Identical graph topologies can have different decentralization profiles when the same functions are assigned to different nodes, so topology alone does not determine decentralization.
  • Systems can be ordered as more or less decentralized by comparing the lengths of their averaged subject vectors, provided the declared subject sets are identical.

Reading between the lines

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

  • Beyond the paper, the framework turns subject selection into a reporting requirement: any decentralization claim becomes comparable only after the declared subjects and weights are stated, which the paper gestures at but does not develop into a standard.
  • The snapshot ontology suggests a natural extension to temporal decentralization, tracking how $\delta(p_u)$ and $\lambda(p_u)$ change across protocol phases, which the paper treats only as separate static topologies.
  • Because the exponential forms of the two metrics are acknowledged as ad hoc, replacing them with alternative functions over the same $r_v$, $r_e$, $\lambda$, and $\delta$ quantities could change comparative orderings; the numbers are not natural constants.
  • The $\delta$-versus-$\lambda$ distinction yields a testable prediction that node-counting proxies miss: two systems with identical node counts can differ in decentralization if one co-locates multiple subject realizations on single vertices.
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

3 major / 4 minor

Summary. The paper argues that 'decentralization' lacks a transferable formal definition, and it proposes a four-layer graph-based ontology to fill this gap. The ontological layer introduces systems, topologies, and subjects; the logical layer defines a subject as centralized or decentralized via the realization multiplicity δ(p_u) of its unique projection particular, and as distributed via the vertex support λ(p_u); the modeling and analytical layers bind these notions to graphs and introduce two metrics, Void Tolerance and Imperviousness, together with a browser-based implementation. The paper validates the framework on federated learning and blockchain instantiations and provides proofs of consistency, disjointness, satisfiability, and entailment for the ontology.

Significance. If the framework worked as claimed, it would provide a subject-relative, graph-grounded separation of decentralization from distribution and a basis for comparing heterogeneous systems. The paper has real strengths: the logical-layer proofs are explicit and structurally clear; the paired instantiations usefully demonstrate that the same topology can host different decentralization profiles; the browser implementation with deterministic simulation is a concrete, reusable artifact; and Section XIV is unusually candid about limitations. However, the central definition depends on an unconstrained analyst-supplied realization count, and the main metric formula has a concrete technical inconsistency. The universal, domain-independent claims are therefore not yet established, although the framework may still be salvageable as a conditional modeling tool.

major comments (3)
  1. [Section IX, CQ1 and Section XII, Eq. (14)] The central dichotomy CENTRALIZED(u) versus DECENTRALIZED(u) reduces to δ(p_u)>1, but δ is not determined by the graph or by the ontology axioms. The axioms in Section IX (Relation to Multiplicity) require only that R_p be finite and non-empty, and Eq. (14) defines δ(p_u)=Σ_v c_u(v), where c_u(v) is a positive integer entered by the user for each vertex. There is no identity criterion for when two realization particulars are distinct. Consequently, the same topology and the same declared subject can be classified either way: the aggregation committee in Instantiation 2 can be modeled as three realizations (δ=3, DECENTRALIZED) or as one joint realization (δ=1, CENTRALIZED) without any change in the graph or the declared subject. Section XIV.D concedes that the metrics are ad hoc but does not address this granularity, and Section XIV.C's claim that δ is the 'furthest context-independent distinction' is unsupported. This is load-bearing for the main contribution, so the paper must either provide an identity criterion for realization particulars or explicitly restrict the definition to a chosen realization model and drop the universal framing.
  2. [Section XI, Proposition 1, Eq. (9)] The displayed formula for Void Tolerance is inconsistent with the stated range T_L∈[0,1] and with the worked examples. Under the stated conditions |G_t|>|G_s| and ε>0, the denominator |G_s|−ε|G_t| is negative for ε=1, so for any r_v>0 the exponent is positive and T_L>1. In the Blockchain example for u_ledger, the text reports T_L=e^{−8/7}≈0.319 with r_v=2, |G_s|=2, and |G_t|=7, which does not follow from the displayed formula as written. Similar concerns apply to the FL examples, where the reported value ≈0.67 for r_v=1, |G_s|=2, and |G_t|=5 also does not match Eq. (9) as typeset. Please correct the formula, check the boundary cases, and recompute all reported analytical values.
  3. [Section XI and Section XII, Eq. (21)] The comparative statements that one system is 'more decentralized' than another depend on the arbitrary choices ε=1 and equal-weight vector averaging. Eq. (9) and Eq. (10) both contain the subject weight ε, and Eq. (21) fixes equal weights; Section XIV.D itself acknowledges that alternative functional forms could produce different numerical scales or comparative orderings. Since Section II (contribution 5) and Section XIII.E present the metrics as enabling direct comparison, the paper should either justify a principled default parameterization or consistently frame the comparative results as illustrative and conditional on the declared weights. As written, the evaluations in Section XI demonstrate internal computational consistency but do not support a general, parameter-independent ordering claim.
minor comments (4)
  1. [Throughout] There are several typos that should be corrected: 'singature' (Section VIII), 'Intantiations' (Section IX), 'decentarlized' (Appendix A1), and '100,000systems' (Section XII).
  2. [Reference list] The reference list is inconsistent: many entries provide volume and pages but omit the publication venue or year (e.g., [26], [27], [29]–[31], [34], [36], [38], [39]), and some entries lack author names (e.g., [14]). Please normalize the bibliography to a single style.
  3. [Section IX, cryptocurrency example] The distinction between logical protocol nodes and physical deployment locations is central to the δ vs λ example, but the paper does not specify how an analyst should choose the vertex abstraction between logical and physical levels. A brief modeling guideline would make the example reproducible.
  4. [Section XIII.D] The sentence that the star-topology vulnerability is not detectable by 'any node-centric metric' is overstated, since classical articulation-point and vertex-cut measures are graph-theoretic and depend on the relational structure; please soften this claim.

Circularity Check

2 steps flagged · score 8.0 of 10

The central classification reduces to analyst-typed realization counts restated as a formal criterion; the 'more decentralized' comparisons are artifacts of arbitrarily chosen metric weights and averaging.

  1. self definitional [Sections IX (Relation to Multiplicity) and XII (Eq. 14)]
    "Projection multiplicity is then the positive-integer-valued function: δ : ΠI → N>0, δ(p)=|Rp| ... For each subject u, the implementation stores a positive realization count cu(v) at every vertex. The projection multiplicity and support cardinality are therefore computed as ... δ(pu)=Σ_{v∈Vu} cu(v)."

    The definition CQ1 makes DECENTRALIZED(u) equivalent to δ(pu)>1, but δ is not derived from the graph or from any identity criterion in the ontology: the axioms only require Rp to be finite and non-empty. In the implementation, δ is the sum of user-entered counts c_u(v), so the formal classification reduces to whether the analyst typed a count greater than 1. The same topology and subject can be flipped between centralized and decentralized by changing c_u(v) or by counting one joint realization versus multiple realizations; e.g., aggregation authority over v1,v2,v3 is counted as three realizations, while nothing rules out one joint realization with δ=1.

  2. other [Section XIV.D; Section XI (Multi-dimensional Aggregate)]
    "Void Tolerance and Imperviousness are purpose-built operators, and although their inputs and boundary conditions follow from the ontology, their particular functional forms are not uniquely entailed by its axioms. ... Alternative functions could encode the same underlying structural quantities while producing different numerical scales or comparative orderings. ... Equal weighting should nevertheless be treated as an explicit aggregation choice rather than as a context-independent feature of decentralization."

    The paper's comparative conclusions (e.g., that the second federated or blockchain example is 'more decentralized') are computed from the equal-weight aggregate A(U)=1/k Σ d_ui with ϵ fixed to 1. The paper itself admits that the functional forms, the sensitivity parameter ϵ, and the aggregation rule are analytical design choices rather than logical consequences of the definition, and that alternative choices can produce different comparative orderings. Consequently, the 'more decentralized' verdict is true by construction of the chosen normalization, not entailed by the ontology. The disclosure in Section XIV.D reduces the pretense but does not remove the circularity: the comparative claim is an artifact of the metric choices, not a consequence of the formal definition of decentralization.

full rationale

The paper's central contribution is a formal definition of decentralization as δ(p_u)>1. The analysis shows this is not a graph-derived property: δ is the cardinality of analyst-declared realization particulars, implemented as user-typed per-vertex counts. Since the axioms impose no identity criterion on realizations, CQ1's classification is true by construction for whatever counts the analyst enters. This is self-definitional circularity in the paper's central claim. The metric layer (Void Tolerance, Imperviousness) has independent graph content for subjects with δ>1, but the comparative aggregate ordering is explicitly admitted to depend on arbitrary functional forms, ϵ, and equal weighting, so the 'more decentralized' conclusions are also chosen, not derived. No load-bearing self-citation was found; the paper's own limitations sections (XIV.D, XIV.E) disclose the arbitrariness but do not resolve it. Since the central definition reduces to the analyst's input, the score is 8.

Assumptions & free parameters 3 free parameters · 5 assumptions · 2 invented entities

The core definition depends on a few assumptions: graph realism, analyst-selected subjects, and cardinality constraints. The metrics add a free parameter ϵ and an arbitrary aggregation rule. The ontology introduces abstract entities, projection and realization particulars, that are not independently observable.

free parameters (3)
  • epsilon (subject weight factor) = 1, uniformly in all evaluations
    Appears in both Void Tolerance and Imperviousness propositions as a strictly positive sensitivity parameter. No assignment procedure is prescribed; fixed to 1 in all examples. It affects how fast metrics reach 0 or 1, and thereby comparative orderings.
  • Reference component selection tie-break = lexicographic by subject support then component order
    Used in Void Tolerance computation (Eq. 15 in the implementation section) to choose which component is the reference when a vertex is removed. This is an arbitrary convention that affects |G_s| and hence the metric value.
  • Aggregation weights = equal weights (1/k) for k subjects
    The multi-dimensional aggregate is computed by simple element-wise averaging. The paper notes this is an explicit aggregation choice, and different weights could yield different orderings.
assumptions (5)
  • domain assumption A system is representable as a non-empty simple undirected graph G=(V,E)
    Ontological commitment to graph-theoretic realism, stated in Section VIII.A. Excludes systems with higher-order or non-pairwise interactions, acknowledged in Section XIV.B.
  • domain assumption Every system has exactly one topology and every subject has exactly one projection particular
    Cardinality axioms in the Logical Layer, Section IX, which are required for the δ and λ functions to be well-defined.
  • domain assumption Each realization particular is located at exactly one vertex, and multiple realizations can co-locate so that 1 ≤ λ(p) ≤ δ(p)
    Defines the relationship between multiplicity and distribution; used in the disjointness proofs and in distinguishing decentralization from distribution.
  • ad hoc to paper Subjects of decentralization are identified by the analyst based on system design intent
    Section XIV.E states that the identification of subjects and the abstraction of real-world systems remain dependent on informed human judgement. This is load-bearing because classifications depend on this choice.
  • domain assumption The chosen graph topology and realization counts c_u(v) faithfully encode the real system
    The modeling layer maps subjects to vertex sets; if the graph is not faithful, all metrics and classifications are void.
invented entities (2)
  • Projection particular (p)
    purpose: Links a system, a topology, and a subject; carries the realization multiplicity δ(p) and support λ(p)
    Introduced as a derived individual in Section VIII.A. Not independently observable; it is a conceptual construct needed to define the metrics.
  • Realization particular (r)
    purpose: Counts distinct instances of a subject; δ(p) = |R_p|
    Abstract entity defined via the hasRealization relation. The distinction between multiple realizations at one vertex is used to separate decentralization from distribution.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Defining Decentralization: An Ontological Perspective." pith.science (2026). https://pith.science/paper/AZ5CE6EJ

@misc{pith2026260809748,
  author       = {Pith},
  title        = {Pith review of: Defining Decentralization: An Ontological Perspective},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AZ5CE6EJ}},
  note         = {Machine review of arXiv:2608.09748}
}
read the original abstract

Decentralization as a concept in computer science has existed for over half a century. Despite its fundamental role across domains such as security, distributed computing, artificial intelligence, cloud infrastructures, and Internet of Things (IoT) architectures, there remains no universally accepted definition of decentralization applicable across computer communication systems. This has become increasingly problematic with the emergence of decentralized AI and machine learning paradigms, including collaborative training, distributed inference, blockchain-based, and agentic AI, where decentralization is often treated as a core design objective. Meanwhile, existing approaches frequently conflate decentralization with related notions such as distribution of trust or specific implementation paradigms. Such ambiguity creates inconsistencies in system analysis, limits comparability between works, and weakens the rigor of formal reasoning surrounding communication architectures and protocol design. In this work, we define this research gap as the Decentralization Problem. We analyze the formal-semantic, epistemological, and pragmatic foundations of decentralization and introduce a graph-based ontology defining it as both relational and subject-specific property of computer communication systems. The framework formally distinguishes decentralization from distribution and supports evaluation through two novel metrics: Void Tolerance and Imperviousness. We also provide a browser-based implementation that enables automated classification and metric computation of arbitrary systems. Instantiations to federated learning and blockchain architectures show consistent, comparable assessments where existing definitions produce incomplete or contradictory conclusions, providing a domain-independent foundation for analysing decentralization across heterogeneous systems.

Figures

Figures reproduced from arXiv: 2608.09748 by the authors.

Figure 1
Figure 1. Ontology Roadmap. systems. A system is taken to exist as a configuration of computational entities and their interconnections, formally representable as a non-empty graph G = (V, E). This commits us to the position that systems are not merely conceptual abstractions but structured objects. The ontology does not address systems that cannot be represented as a graph, thereby implicitly constraining the domain to compu… view at source ↗
Figure 2
Figure 2. System-class taxonomic backbone. hasTopology Topology hasSubject ofSubject Subject hasProjection System is_a Data is_a Model Ownership . . . encodes Adjacency Matrix encodes Incidence Matrix . . . inTopology hasRealization p realizedAt r v [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. System-centric ontology schema. Individuals p, r, and v denote [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Star-topology centralized federated learning setup with one aggregator [PITH_FULL_IMAGE:figures/full_fig_p018_4.png]
Figure 5
Figure 5. Figure 5: Decentralized federated learning setup with three aggregation author [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]
Figure 6
Figure 6. Figure 6: Shared blockchain topology used for both blockchain examples. [PITH_FULL_IMAGE:figures/full_fig_p019_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 39 canonical work pages

  1. [1]

    On distributed communications: I. introduction to distributed communications networks

    P. Baran, “On distributed communications: I. introduction to distributed communications networks.” [Online]. Available: https://www.rand.org/ pubs/research_memoranda/RM3420.html

    work page

  2. [2]

    Communication-efficient learning of deep networks from decentralized data,

    H. B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, “Communication-efficient learning of deep networks from decentralized data,” 2016. [Online]. Available: https://arxiv.org/abs/1602.05629

    arXiv 2016

  3. [3]

    Decentralized federated learning: A survey and perspective,

    L. Yuan, Z. Wang, L. Sun, P. S. Yu, and C. G. Brinton, “Decentralized federated learning: A survey and perspective,”IEEE Internet of Things Journal, vol. 11, no. 21, pp. 34 617–34 638, 2024

    2024

  4. [4]

    Decentralized federated learning: Fundamentals, state of the art, frameworks, trends, and challenges,

    E. T. Martínez Beltrán, M. Q. Pérez, P. M. S. Sánchez, S. L. Bernal, G. Bovet, M. G. Pérez, G. M. Pérez, and A. H. Celdrán, “Decentralized federated learning: Fundamentals, state of the art, frameworks, trends, and challenges,”IEEE Communications Surveys & Tutorials, vol. 25, no. 4, pp. 2983–3013, 2023

    work page 2023

  5. [5]

    Swarm intelligence decentralized decision making in multi-agent system,

    A. A. Joseph, G. S. Nambiar, and N. Jayapandian, “Swarm intelligence decentralized decision making in multi-agent system,” in2023 8th International Conference on Communication and Electronics Systems (ICCES), 2023, pp. 1425–1430

    work page 2023

  6. [6]

    Verillm: A lightweight framework for publicly verifiable decentralized inference,

    K. Wang, Z. Zhao, X. Song, Z. Li, L. Xia, C. Tong, B. Shi, W. Qu, E. Yang, and L. Ai, “Verillm: A lightweight framework for publicly verifiable decentralized inference,” 2026. [Online]. Available: https://arxiv.org/abs/2509.24257

    arXiv 2026

  7. [7]

    Blockchain for ai: Review and open research challenges,

    K. Salah, M. H. U. Rehman, N. Nizamuddin, and A. Al-Fuqaha, “Blockchain for ai: Review and open research challenges,”IEEE Access, vol. 7, pp. 10 127–10 149, 2019

    work page 2019

  8. [8]

    A comprehensive review on understanding the decentralized and collaborative approach in machine learning,

    S. Saif, M. J. Islam, M. Z. B. Jahangir, P. Biswas, A. Rashid, M. A. A. Nasim, and K. D. Gupta, “A comprehensive review on understanding the decentralized and collaborative approach in machine learning,”

    work page

Show all 56 references
  1. [9]

    Sok: Blockchain-based decentralized ai (deai),

    E. Lui, R. Sun, V . Shah, X. Xiong, J. Sun, D. Crapis, W. Knottenbelt, and Z. Wang, “Sok: Blockchain-based decentralized ai (deai),” 2026. [Online]. Available: https://arxiv.org/abs/2411.17461

    2026

  2. [10]

    Decentralisation: A multidisciplinary perspective,

    B. Bodó, J. K. Brekke, and J.-H. Hoepman, “Decentralisation: A multidisciplinary perspective,”Internet Policy Review, vol. 10, no. 2, pp. 1–21, 2021. [Online]. Available: https://hdl.handle.net/10419/235966

    2021

  3. [11]

    The byzantine generals problem,

    L. Lamport, R. Shostak, and M. Pease, “The byzantine generals problem,” vol. 4, no. 3, pp. 382–401. [Online]. Available: https: //dl.acm.org/doi/10.1145/357172.357176

  4. [12]

    Towards robust distributed systems,

    E. A. Brewer, “Towards robust distributed systems,” inProceedings of the nineteenth annual ACM symposium on Principles of distributed computing, ser. PODC ’00. Association for Computing Machinery, p. 7. [Online]. Available: https://dl.acm.org/doi/10.1145/343477.343502

  5. [13]

    Bitcoin: A peer-to-peer electronic cash system,

    S. Nakamoto, “Bitcoin: A peer-to-peer electronic cash system,” 2008

    2008

  6. [14]

    B. S. Srinivasan. Quantifying decentralization. [Online]. Available: https://news.earn.com/quantifying-decentralization-e39db233c28e

  7. [15]

    Impossibility of full decentralization in permissionless blockchains,

    Y . Kwon, J. Liu, M. Kim, D. Song, and Y . Kim, “Impossibility of full decentralization in permissionless blockchains,” inProceedings of the 1st ACM Conference on Advances in Financial Technologies, ser. AFT ’19. Association for Computing Machinery, 2019, pp. 110–123. [Online]...

    2019

  8. [16]

    Measuring decentralization in bitcoin and ethereum using multiple metrics and granularities,

    Q. Lin, C. Li, X. Zhao, and X. Chen, “Measuring decentralization in bitcoin and ethereum using multiple metrics and granularities,” in2021 IEEE 37th International Conference on Data Engineering Workshops (ICDEW), pp. 80–87, ISSN: 2473-3490. [Online]. Available: https://ieeexpl...

  9. [17]

    Towards a theory of digital network de/centralization: Platform-infrastructure lessons drawn from blockchain

    E. Rossi and C. Sørensen, “Towards a theory of digital network de/centralization: Platform-infrastructure lessons drawn from blockchain.” [Online]. Available: https://papers.ssrn.com/abstract= 3503609

  10. [18]

    Systematizing decentralization and privacy: Lessons from 15 years of research and deployments,

    C. Troncoso, M. Isaakidis, G. Danezis, and H. Halpin, “Systematizing decentralization and privacy: Lessons from 15 years of research and deployments,”Proceedings on Privacy Enhancing Technologies, vol. 2017, no. 4, p. 404–426, Oct. 2017. [Online]. Available: http://dx.doi.org/...

    2017 doi

  11. [19]

    Toward quantifying decentralization of blockchain networks with relay nodes,

    Y . Shahsavari, K. Zhang, and C. Talhi, “Toward quantifying decentralization of blockchain networks with relay nodes,” Frontiers in Blockchain, vol. V olume 5 - 2022, 2022. [Online]. Available: https://www.frontiersin.org/journals/blockchain/ articles/10.3389/fbloc.2022.812957

    2022

  12. [20]

    Decentralization in bitcoin and ethereum networks,

    A. E. Gencer, S. Basu, I. Eyal, R. van Renesse, and E. G. Sirer, “Decentralization in bitcoin and ethereum networks,” inFinancial Cryptography and Data Security: 22nd International Conference, FC 2018, Nieuwpoort, Curaçao, February 26 – March 2, 2018, Revised Selected Papers. ...

    2018 doi

  13. [21]

    A typology of multidimensional adjectives,

    G. W. Sassoon, “A typology of multidimensional adjectives,”Journal of Semantics, vol. 30, no. 3, pp. 335–380, 08 2013. [Online]. Available: https://doi.org/10.1093/jos/ffs012

    2013 doi

  14. [22]

    Über formal unentscheidbare sätze der principia mathematica und verwandter systeme i,

    K. Gödel, “Über formal unentscheidbare sätze der principia mathematica und verwandter systeme i,” vol. 38, no. 1, pp. 173–198. [Online]. Available: https://doi.org/10.1007/BF01700692

    doi

  15. [23]

    Toulmin,The Uses of Argument

    S. Toulmin,The Uses of Argument. Cambridge, England: Cambridge University Press, 1958

    1958

  16. [24]

    M. A. Finocchiaro,Informal Logic and the Theory of Reasoning (1984). Cambridge University Press, 2005, p. 21–33

    1984

  17. [25]

    New York, NY: Cambridge University Press, 2005

    ——,Arguments About Arguments: Systematic, Critical, and Historical Essays in Logical Theory. New York, NY: Cambridge University Press, 2005. [Online]. Available: https://assets.cambridge.org/97805216/ 18533/frontmatter/9780521618533_frontmatter.pdf

    2005

  18. [26]

    Centralized versus decentralized computing: organizational considerations and management options,

    J. L. King, “Centralized versus decentralized computing: organizational considerations and management options,” vol. 15, no. 4, pp. 319–349. [Online]. Available: https://dl.acm.org/doi/10.1145/289.290

    doi

  19. [27]

    Data security in decentralized cloud systems – system comparison, requirements analysis and organizational levels,

    A. Müller, A. Ludwig, and B. Franczyk, “Data security in decentralized cloud systems – system comparison, requirements analysis and organizational levels,” vol. 6, no. 1, p. 15. [Online]. Available: https://doi.org/10.1186/s13677-017-0082-3

    doi

  20. [28]

    The concept of decentralization through time and disciplines: a quantitative exploration,

    G. Di Bona, A. Bracci, N. Perra, V . Latora, and A. Baronchelli, “The concept of decentralization through time and disciplines: a quantitative exploration,” vol. 12, no. 1, p. 42. [Online]. Available: https://doi.org/10.1140/epjds/s13688-023-00418-1

    doi

  21. [29]

    Cross-referencing method for scalable public blockchain,

    T. Yanagihara and A. Fujihara, “Cross-referencing method for scalable public blockchain,” vol. 15, p. 100419. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S2542660521000639

  22. [30]

    Duopoly competition in blockchain game with interoperability,

    J. Sun, H. Zhang, A. El Saddik, and W. Cai, “Duopoly competition in blockchain game with interoperability,” in2024 IEEE International Conference on Systems, Man, and Cybernetics (SMC), pp. 4881–

  23. [31]

    Token incentive and blockchain developer contribution,

    N. Luo and Z. Zheng, “Token incentive and blockchain developer contribution,”Information Systems Research, 2026. [Online]. Available: https://pubsonline.informs.org/doi/abs/10.1287/isre.2023.0480

    2026

  24. [32]

    Towards resilient internet of things: Vision, challenges, and research roadmap,

    C. Tsigkanos, S. Nastic, and S. Dustdar, “Towards resilient internet of things: Vision, challenges, and research roadmap,” in2019 IEEE 39th International Conference on Distributed Computing Systems (ICDCS), 2019, pp. 1754–1764

    2019

  25. [33]

    Blockchain security and privacy: Threats, challenges, applications, and tools,

    A. Alghuried, M. Alkinoon, M. Mohaisen, A. Wang, C. Zou, and D. Mohaisen, “Blockchain security and privacy: Threats, challenges, applications, and tools,”Distrib. Ledger Technol., vol. 5, no. 1, Dec

  26. [34]

    Ontology modeling for decentralized household energy systems,

    J. Wu, F. Orlandi, T. AlSkaif, D. O’Sullivan, and S. Dev, “Ontology modeling for decentralized household energy systems,” in2021 International Conference on Smart Energy Systems and Technologies (SEST), pp. 1–6. [Online]. Available: https://ieeexplore.ieee.org/abstract/ docume...

  27. [35]

    DECENT: An ontology for decentralized governance in the renewable energy sector,

    F. Kaya and J. Gordijn, “DECENT: An ontology for decentralized governance in the renewable energy sector,” in2021 IEEE 23rd Conference on Business Informatics (CBI), vol. 01, pp. 11– 20, ISSN: 2378-1971. [Online]. Available: https://ieeexplore.ieee.org/ abstract/document/9610686

    1971

  28. [36]

    Available: https://doi.org/10.1145/3716323

    [Online]. Available: https://doi.org/10.1145/3716323

    doi

  29. [37]

    Ontology-based approach to decentralized production control in the context of cloud manufacturing execution systems

    B. Katti, “Ontology-based approach to decentralized production control in the context of cloud manufacturing execution systems.” [Online]. Available: https://kluedo.ub.rptu.de/frontdoor/index/index/docId/5987

  30. [38]

    Adaptive information sharing with ontological relevance computation for decentralized self-organization systems,

    W. Liu, W. Ran, S. Nantogma, and Y . Xu, “Adaptive information sharing with ontological relevance computation for decentralized self-organization systems,” vol. 23, no. 3, p. 342. [Online]. Available: https://www.mdpi.com/1099-4300/23/3/342

  31. [39]

    An ontology framework towards decentralized information management for eco- industrial parks,

    L. Zhou, C. Zhang, I. A. Karimi, and M. Kraft, “An ontology framework towards decentralized information management for eco- industrial parks,” vol. 118, pp. 49–63. [Online]. Available: https: //www.sciencedirect.com/science/article/pii/S0098135418300929

  32. [40]

    The decentralization of knowledge,

    H. Halpin and A. Monnin, “The decentralization of knowledge,” vol. 21, no. 12. [Online]. Available: https://hal.science/hal-01397931

  33. [41]

    Formal ontology and information systems,

    N. Guarino, “Formal ontology and information systems,” inProceedings of the First International Conference (FOIS’98), vol. 46. Trento, Italy: IOS Press, 1998, pp. 3–15. [Online]. Available: https: //ai.ia.agh.edu.pl/wiki/_media/pl:dydaktyka:semweb:guarinofois98.pdf

    1998

  34. [42]

    Toward a formal scholarly understanding of blockchain-mediated decentralization: A systematic review and a framework,

    M. R. Hoffman, L.-D. Ibáñez, and E. Simperl, “Toward a formal scholarly understanding of blockchain-mediated decentralization: A systematic review and a framework,” vol. 3. [Online]. Available: https://www.frontiersin.org/journals/blockchain/articles/10. 3389/fbloc.2020.00035/full

  35. [43]

    The descrip- tion logic handbook: Theory, implementation and applica- tions

    F. Baader, D. Calvanese, D. L. McGuinness, D. Nardi, and P. F. Patel-Schneider, “The descrip- tion logic handbook: Theory, implementation and applica- tions.” [Online]. Available: https://www.cambridge.org/core/books/ description-logic-handbook/F050683766E57EE9BB07BC01BB7A7069

  36. [44]

    R. Arp, B. Smith, and A. D. Spear,Building Ontologies with Basic Formal Ontology. The MIT Press, 2015. [Online]. Available: http://www.jstor.org/stable/j.ctt17kk7vw

    2015

  37. [45]

    Ontology development 101: A guide to creating your first ontology,

    N. F. Noy and D. L. McGuinness, “Ontology development 101: A guide to creating your first ontology,” 2004. [Online]. Available: https://protege.stanford.edu/publications/ontology_ development/ontology101-noy-mcguinness.html

    2004

  38. [46]

    Methodology for the design and evaluation of ontologies,

    M. Gruninger and M. S. Fox, “Methodology for the design and evaluation of ontologies,”Proc. IJCAI’95, Workshop on Basic Ontological Issues in Knowledge Sharing, 1995. [Online]. Available: https://cir.nii.ac.jp/crid/1571135650978573440

    1995

  39. [47]

    Advances and open problems in federated learning,

    P. Kairouz, H. B. McMahan, B. Avent, A. Bellet, M. Bennis, A. N. Bhagoji, K. Bonawitz, Z. Charles, G. Cormode, R. Cummings, R. G. L. D’Oliveira, H. Eichner, S. E. Rouayheb, D. Evans, J. Gardner, Z. Garrett, A. Gascón, B. Ghazi, P. B. Gibbons, M. Gruteser, Z. Harchaoui, C. He, ...

    2021 arXiv

  40. [48]

    Secure aggregation for federated learning in flower,

    K. H. Li, P. P. B. de Gusmão, D. J. Beutel, and N. D. Lane, “Secure aggregation for federated learning in flower,” 2022. [Online]. Available: https://arxiv.org/abs/2205.06117

    2022 arXiv

  41. [49]

    Deep reinforcement learning-based scheduling in distributed systems: A critical review,

    Z. Jalali Khalil Abadi, N. Mansouri, and M. M. Javidi, “Deep reinforcement learning-based scheduling in distributed systems: A critical review,” vol. 66, no. 10, pp. 5709–5782. [Online]. Available: https://doi.org/10.1007/s10115-024-02167-7

    doi

  42. [50]

    Veríssimo and L

    P. Veríssimo and L. Rodrigues,Distributed Systems for System Archi- tects. Springer Science & Business Media, 2012

    2012

  43. [51]

    Efficient and less centralized federated learning,

    L. Chou, Z. Liu, Z. Wang, and A. Shrivastava, “Efficient and less centralized federated learning,” inMachine Learning and Knowledge Discovery in Databases. Research Track, N. Oliver, F. Pérez-Cruz, S. Kramer, J. Read, and J. A. Lozano, Eds. Springer International Publishing, p...

  44. [52]

    Decenntralization ontology sandbox,

    Anonymous, “Decenntralization ontology sandbox,” Aug 2026. [Online]. Available: https://szelagjk.github.io/DecentralizationOntology/

    2026

  45. [53]

    Depth-first search and linear graph algorithms,

    R. Tarjan, “Depth-first search and linear graph algorithms,”SIAM Journal on Computing, vol. 1, no. 2, pp. 146–160, 1972. [Online]. Available: https://doi.org/10.1137/0201010 APPENDIX A1. Aristotelian DefinitionsHaving created our taxonomy, we establish Aristotelian definitions...

    1972 doi

  46. [54]

    Multidimensional adjectives,

    J. D’Ambrosio and B. Hedden, “Multidimensional adjectives,” Australasian Journal of Philosophy, vol. 102, no. 2, pp. 253–277, 2024. [Online]. Available: https://doi.org/10.1080/00048402.2023.2277923

    2024

  47. [2025]

    Available: https://arxiv.org/abs/2503.09833

    [Online]. Available: https://arxiv.org/abs/2503.09833

    arXiv

  48. [4886]

    Available: https://ieeexplore.ieee.org/abstract/document/ 10831351

    [Online]. Available: https://ieeexplore.ieee.org/abstract/document/ 10831351

Pith tools

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