Pith. sign in

REVIEW 4 major objections 5 minor 60 references

Three Kinds of Negation in Knowledge and Their Mathematical Foundations

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper argues that negation in knowledge is three distinct operations—contradictory, opposite, and intermediary—and it builds a fuzzy set theory and a three-valued logic in which all three can be represented and reasoned about.

desk verdict The philosophical taxonomy is readable, but the logic's central soundness claim collapses on the paper's own three-valued table. read the letter →

arxiv 2505.24422 v1 pith:L5PEXEFA submitted 2025-05-30 cs.AI

Zhenghua Pan , Yong Wang This is my paper · ORCID
classification cs.AI MSC 03B5003B5203B6003E72
keywords negationinknowledgecontradictoryoppositeintermediarythree-valuedlogicfuzzysetssoundnessandcompletenessrepresentation
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 aims to establish that the single word “negation” hides three logically different relations in knowledge: contradictory negation, opposite negation, and intermediary negation. It argues that the classical distinction between contradiction and opposition, plus the mediating term between opposites, forces this three-way split, and that a knowledge-representation formalism should treat the three as separate operations. To that end it constructs SCOI, a fuzzy-set algebra whose membership functions close under the three negations, and LCOI, an axiomatic propositional logic with the same three negations under a three-valued semantics. The paper’s central formal claim is that LCOI is sound and complete for that semantics, and that the derived laws of non-contradiction hold for the pairwise combinations among a formula, its opposite, and its intermediary. A sympathetic reader would care because, if the framework is right, “neither P nor its opposite” sentences—common in clear and fuzzy knowledge alike—can be represented and reasoned about as first-class objects rather than awkward conjunctions.

What carries the argument

The load-bearing object is the triple of negation operators tied by the identity $\neg A = \bar{A} \lor \tilde{A}$ at the level of truth values, together with the companion identity $\tilde{A} = \neg A \land \neg\bar{A}$, which expresses intermediary negation as “not A and not the opposite of A.” The threshold parameter $\lambda \in (0,1)$ in the piecewise definition of intermediary negation is what gives the third negation its content: it partitions $[0,1]$ into high, medium, and low zones, and $\tilde{A}$ maps values from the outer zones into the middle zone. The three-valued assignment interprets $\neg$ as the maximum of the other two negations, with implication given by a three-valued table, so the whole system rests on these operator identities and the threshold geometry. The completeness argument then follows the standard route of maximal consistent sets and a canonical valuation; the truth lemma for that valuation is the step that connects syntactic consistency to semantic satisfaction.

What would settle it

A concrete check: evaluate each axiom of LCOI, especially axioms (a12) and (a13), under the three-valued tables in Definition 2 of Section 4.2.2; if any axiom takes value $\frac12$ under some assignment, the soundness theorem is false and the stated semantics is not a semantics for LCOI. Conversely, if a formula receives value 1 under every three-valued assignment yet is not derivable, the completeness theorem is false.

Watch

Extended reading notes

Core claim

The paper’s claim is that the traditional single negation operator should be replaced by three connected operators. On sets, for any membership degree $A(x)$, opposite negation is the usual complement $A^{\bar{}}(x) = 1 - A(x)$; intermediary negation is a piecewise function $A^{\tilde{}}(x)$ that sends values near the extremes into a middle band controlled by a parameter $\lambda$; and contradictory negation is the maximum of the two, $A^{\neg}(x) = \max(A^{\bar{}}(x), A^{\tilde{}}(x))$. In the logic LCOI the same three appear as connectives, with contradictory negation defined by $\neg A = \bar{A} \lor \tilde{A}$, and the semantics is three-valued with values $0, \frac12, 1$. The paper proves operational properties of SCOI (monotonicity, De Morgan-style laws for opposite negation, and classification of the three as strict or strong fuzzy negations) and derives in LCOI a battery of inference theorems, including three new laws of non-contradiction. Its headline formal assertion is that LCOI is sound and complete with respect to the three-valued semantics.

Load-bearing premise

The soundness-and-completeness claim rests on the assumption that the truth values assigned by the canonical valuation on atomic formulas extend to all formulas so that the truth lemma holds for every formula; if that extension fails for the negation or implication connectives, the completeness direction does not follow.

Editorial extensions

If this is right

  • If LCOI is sound, every theorem proved in Section 4.2 is guaranteed true under the three-valued semantics; in particular the three new laws of non-contradiction—$\vdash \neg(A \land \tilde{A})$, $\vdash \neg(\bar{A} \land \tilde{A})$, and $\vdash \neg(A \land \bar{A})$—are semantically valid.
  • Intermediary negation becomes an expressible, computable concept: a statement like “x is neither a positive nor a negative integer” reduces to the single atomic formula $\tilde{A}(x)$ rather than a conjunction of unrelated negations.
  • The threshold $\lambda$ gives a parameterized notion of “middle,” so the same formal framework can be tuned to different contexts by moving the boundary between high, medium, and low membership.
  • In the decision-making application, the three negations change which fuzzy production rules fire: the example recommends scheme (d) because the intermediate membership zones, rather than the high or low zones, are the ones that satisfy the thresholds.

Reading between the lines

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

  • A natural extension the paper does not explore is to treat $\lambda$ as a variable of the semantics, yielding a parametrized family of three-valued logics whose theorems may shift as the middle band widens or narrows.
  • One testable consequence of the identity $\tilde{A} = \neg A \land \neg\bar{A}$ is that the algebra of iterated negations is small; computing the full table of composites of $\neg$, $\bar{}$, and $\tilde{}$ would show whether every formula collapses to one of a few normal forms.
  • Comparing the set of three-valued valid formulas of LCOI with those of standard strong-Kleene or Łukasiewicz logic would locate this system among known three-valued logics and could reveal which theorems depend specifically on the asymmetry of the implication table.
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

4 major / 5 minor

Summary. The paper argues that knowledge contains three distinct kinds of negation — contradictory, opposite, and intermediary — and proposes two formal frameworks: SCOI, a fuzzy-set structure with three negation operations, and LCOI, a propositional logic with the corresponding negations. The central formal claim, stated in the abstract and in Section 4.2.2, is that LCOI is both sound and complete under a three-valued semantic interpretation. The paper also derives many formal theorems for LCOI and illustrates the framework on a multi-attribute decision-making example.

Significance. If the formal results were correct, the paper would provide a useful mathematical treatment of three negations and a logic tailored to knowledge representation problems that distinguish contradiction from opposition. The philosophical taxonomy and the broad survey of negation in various fields are competently assembled, and the application example shows the intended use of the framework. However, the load-bearing formal claims fail: at least two LCOI axioms are invalid under the paper's own semantics, and the completeness proof relies on a truth lemma whose induction is not valid. These are not presentation issues; they invalidate the headline 'sound and complete' claim. The framework may still be repairable, but the current manuscript does not establish its main result.

major comments (4)
  1. [Section 4.2.2, Definition 2 and Theorem 1] Axiom (a2), (A→(A→B))→(A→B), is not valid under the three-valued implication table given in Definition 2. With v(A)=1/2 and v(B)=0, the table gives v(A→B)=1/2, v(A→(A→B))=1/2→1/2=1, and then v((A→(A→B))→(A→B))=1→1/2=1/2. Similarly, axiom (a6), ¬_c A→(A→B), fails: with v(A)=1/2 and v(B)=0, Definition 2(5) gives v(¬_c A)=max(v(╕A),v(~A))=max(1/2,1)=1, so v(¬_c A→(A→B))=1→1/2=1/2. Thus both axioms are counterexamples to the claim in the proof of Theorem 1 that 'the axioms in LCOI are all valid formulas.' The soundness theorem is therefore false, and the abstract's assertion that LCOI is sound and complete is not supported.
  2. [Section 4.1, Definition 1 and Section 4.1.2, Proposition 5] Proposition 5 claims that the intermediary negation φ~ is a strict fuzzy negation, but the proof does not establish the required condition (N1), namely φ~(1)=0 and φ~(0)=1. The proof shows that in case (a), φ~(1)=1−λ, and in case (d), φ~(1)=λ, and then argues by taking limits as λ→1 or λ→0. For any fixed λ∈(0,1), however, these values are not 0; similarly, φ~(0) is not 1. Hence the function defined in Definition 1 does not satisfy Definition 4(N1) for the actual λ used in the construction. The assertion that the intermediary negation is a strict fuzzy negation is therefore not proven and is in fact false for the given definition.
  3. [Section 4.2.2, Lemma 5] The truth lemma is not established for composite formulas. Definition 5 defines the valuation only on atomic formulas p, by membership of p, ╕p, or ~p in a maximal consistent set. In the induction step of Lemma 5, the case A=╕B uses the inference 'v(B)=0, so ╕B∈Σ by Definition 5,' and the case A=~B repeatedly invokes Definition 5 on formulas such as ~B and ╕~B. Since Definition 5 has no clauses for negated formulas, these steps are invalid. Consequently, the induction does not prove Lemma 5 for formulas containing negations or implications, and the completeness theorem (Theorem 2), which relies on Lemma 5, does not follow from the given proof.
  4. [Section 4.2.2, Lemma 3(1) and Lemma 4(1)] Lemma 3(1) assumes that for any formula A, if A∉Σ then either ╕A∈Σ or ~A∈Σ. This trichotomy is not justified by Lemma 4(1), which only proves that at most one of A, ╕A, ~A belongs to a maximal consistent set; it does not prove that at least one belongs. The proof of Lemma 4(1) only rules out the presence of two of them, using Lemma 2. Without a proof of the trichotomy, the canonical valuation in Definition 5 is not well-defined for all atomic formulas, and the completeness argument is incomplete even before the composite-formula issue noted above.
minor comments (5)
  1. [Section 4.3] The symbols for opposite and intermediary negation are interchanged in the sentence 'For the contradictory negation A>, opposite negation A~ and intermediary negation A╕ of the set A'; according to Definition 1, A╕ is the opposite negation and A~ is the intermediary negation.
  2. [Section 4.2] The paper uses two separate numbering systems for definitions and theorems: Definition 1 and Theorem 1 appear in both Section 4.2.1 and Section 4.2.2, which makes cross-references ambiguous.
  3. [Section 4.2.1, Theorem 1 proof [4]] The displayed formula in step (3) of the proof has unbalanced parentheses and does not appear to be a well-formed formula; it should be corrected or rewritten.
  4. [Section 4.1.2, Proposition 5 proof] The sentence 'by the proposition 3, φ╕ is a decreasing function' should refer to φ~; Proposition 3 asserts that both φ╕ and φ~ are decreasing, so this is a typographical error.
  5. [Section 5] The symbol λ is used both for the SCOI parameter in Definition 1 of Section 4.1 and for the threshold values computed in Table 4; the relationship between these two uses is never explained, which may confuse readers.

Circularity Check

0 steps flagged · score 0.0 of 10

No constructional circularity: the SCOI/LCOI formal framework is defined rather than derived from its own theorems, and the soundness/completeness failure is a correctness gap, not a reduction to inputs.

full rationale

SCOI's three negations are introduced by Definition 1 (Section 4.1) as primitive construction rules, and the subsequent propositions and properties are derived from that definition. The piecewise definition of the intermediary negation is a definitional choice, not a fitted prediction smuggled in as a result. LCOI's semantics (Section 4.2.2, Definition 2) is a recursive truth assignment; the syntactic abbreviation 'A = ╕A∨~A' is mirrored by semantic clause (5), but this is a normal definitional correspondence to SCOI Property 3(3), not a claim that the semantics is discovered from the axioms. The self-citations to Pan et al. supply the prior framework, yet the needed definitions are restated in full, so no load-bearing conclusion reduces to an unverified self-citation. The serious problem in the paper is not circularity: Theorem 1(i) asserts 'Since the axioms in LCOI are all valid formulas' without verifying the axioms, and under the paper's own three-valued table (a2) and (a6) fail at v(A)=1/2, v(B)=0, so the claimed soundness theorem is false as stated; Lemma 5 also applies an atomic-formula definition to composite formulas. These are proof and correctness failures, not reductions of a prediction to its fitted input. The decision-making example estimates thresholds from survey data and then applies them to a new individual; it is an illustrative computation, not a renamed prediction. No circular step meeting the quoted-equation standard was found.

Assumptions & free parameters 3 free parameters · 6 assumptions · 1 invented entities

The ledger counts the parameters and assumptions the central claims depend on. The threshold lambda and the provincial thresholds are fitted to the same data used to validate the decision example. The semantic tables and axioms of LCOI are stipulated and are mutually inconsistent for at least two axioms. The only invented conceptual entity is intermediary negation, which is defined to match the examples rather than measured independently.

free parameters (3)
  • lambda (intermediary-negation threshold) = 0.819 for income, 0.813 for savings in Section 5, Table 4
    Introduced in Definition 1 (Section 4.1) as a variable parameter that controls the range of membership values for the intermediary negation. In the decision example, lambda is fitted by averaging a computed membership value with 1, then averaging across provinces, and it determines the final recommendation.
  • Province-level thresholds alpha = 1.000, 0.888, 0.655, 0.732 (income); 1.000, 0.879, 0.637, 0.734 (savings) in Table 3
    Each provincial threshold is the average of a distance-ratio membership score and the assumed value 1; these are then averaged into the global lambda. These are fit values, not derived constraints.
  • Elasticity values epsilon = +/-500 (high income), +/-100 (low income), +/-20000 (high deposit), +/-10000 (low deposit)
    Hand-set in Section 5 Step 1 to 'improve accuracy' of integrated survey data; they shift the membership function boundaries and influence all subsequent thresholds.
assumptions (6)
  • standard math Zadeh's definition of fuzzy sets (membership functions to [0,1])
    Invoked at the start of Section 4.1, Definition 1; a standard background theory.
  • domain assumption Contradiction and opposition exhaust the kinds of relation between species under a genus, with opposition the extreme form of contradiction
    Section 3.2 adopts this from formal logic and Hegelian dialectics; the paper's entire taxonomy depends on it.
  • domain assumption All opposite concepts transition through intermediaries
    Section 3.2, point b, citing Hegel and Engels; used to justify treating the 'middle' as a genuine negation.
  • ad hoc to paper The three-valued semantics for LCOI (Definition 2, Section 4.2.2)
    The truth tables are stipulated to support the desired connectives; however axioms (a2) and (a6) are not valid in them, so the stipulated semantics does not fit the axiom system.
  • ad hoc to paper Axiom schemas (a1)-(a13) of LCOI
    The axioms are chosen by the authors; the paper claims they are valid in the semantics but does not verify, and two are actually invalid.
  • domain assumption Distance Ratio Function with Euclidean distance (Zhang and Pan 2009) for membership computation
    Used in Section 5 Step 3 to define membership degrees from survey data; comes from the same research group, its appropriateness is not independently justified.
invented entities (1)
  • Intermediary negation (operator ~ on formulas and phi~ on membership values)
    purpose: To represent the in-between 'neither...nor...' state (e.g., zero between positive and negative integers, dusk between day and night) as a third kind of negation.
    phi~ is defined piecewise with a fitted parameter lambda and is tuned so that the desired ordering and boundary properties appear; the paper's examples (zero, dusk, moderate income) are used both to motivate the definition and to illustrate it, so there is no falsifiable handle outside the paper. The claim that it 'exists in objective reality' is asserted, not demonstrated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Three Kinds of Negation in Knowledge and Their Mathematical Foundations." pith.science (2026). https://pith.science/paper/L5PEXEFA

@misc{pith2026250524422,
  author       = {Pith},
  title        = {Pith review of: Three Kinds of Negation in Knowledge and Their Mathematical Foundations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L5PEXEFA}},
  note         = {Machine review of arXiv:2505.24422}
}
read the original abstract

In the field of artificial intelligence, understanding, distinguishing, expressing, and computing the negation in knowledge is a fundamental issue in knowledge processing and research. In this paper, we examine and analyze the understanding and characteristics of negation in various fields such as philosophy, logic, and linguistics etc. Based on the distinction between the concepts of contradiction and opposition, we propose that there are three different types of negation in knowledge from a conceptual perspective: contradictory negation, opposite negation, and intermediary negation. To establish a mathematical foundation that fully reflects the intrinsic connections, properties, and laws of these different forms of negation, we introduce SCOI: sets with contradictory negation, opposite negation and intermediary negation, and LCOI: logic with contradictory negation, opposite negation and intermediary negation, and we proved the main operational properties of SCOI as well as the formal inference relations in LCOI.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 55 canonical work pages

  1. [1]

    (2014).1-Computational Framework, Editor(s): Syed V

    Ahamed, S.V . (2014).1-Computational Framework, Editor(s): Syed V . Ahamed, Next Generation Knowledge Machines, Elsevier, 3–36. https://doi.org/10.1016/B978-0-12-416629-5.00001-3

    work page doi:10.1016/b978-0-12-416629-5.00001-3 2014

  2. [2]

    P., et al

    Arnaout, H., Nguyen, T. P., et al. (2023). UnCommonSense in Action! Informative Negations for Commonsense Knowledge Bases. Proc. of the 16th ACM International WSDM Conference, February 27-March 3, Singapore, Singapore. pp. 1120−1123. https://doi.org/10.1145/3539597.3573027

    arXiv 2023

  3. [3]

    Simon, R., Gerhard, W., et al

    Arnaout, H. Simon, R., Gerhard, W., et al. (2022). UnCommonSense: Informative Negative Knowledge about Everyday Concepts. Proceedings of the 31st ACM Int’l Conference on Information and Knowledge Management (CIKM ’22), Atlanta, GA, USA. ACM, Oct. pp.17–21. https://doi.org/10.1145/3511808.3557484

    arXiv 2022

  4. [4]

    Anjaria, K. (2020). Negation and entropy: Effectual knowledge management equipment for learning organizations. Expert Systems with Applications, 157, 113497

    work page 2020

  5. [5]

    and Wagner, G

    Analyti, A., Antoniou, G., Damasio, C.V . and Wagner, G.. (2004). Negation and Negative Information in the W3C Resource Description Framework. Annals of Mathematics, Computing & Teleinformatics (AMCT), 1(2): 25–34

    work page 2004

  6. [6]

    Ayer, A. J. (1952). Negation, Journal of Philosophy, 49:797–815

    work page 1952

  7. [7]

    Beeson, M., Veroff, R., Wos, L. (2005). Double-Negation Elimination in Some Propositional Logics. Studia Logica, 80: 195–234

    work page 2005

  8. [8]

    Carter, J. A. (2023). Digital Knowledge: A Philosophical Investigation.Taylor and Francis

    work page 2023

Show all 60 references
  1. [9]

    P, Taboada, M., Mitkov, R

    Cruz, N. P, Taboada, M., Mitkov, R. (2016). A machine-learning approach to negation and speculation detection for sentiment analysis. Journal of the Association for Information Science and Technology, 67(9): 2118–2136. https://doi.org/10.1002/asi.23533

    2016 doi

  2. [10]

    Chen, L. (2019). Attention-based deep learning system for negation and assertion detection in clinical notes. International Journal of Artificial Intelligence and Applications, 10(1): 1–9

    2019

  3. [11]

    Dimiter, V. (2005). Nelson’s Negation on the Base of Weaker Versions of Intuitionistic Negation. Studia Logica, 80: 393–430

    2005

  4. [12]

    M, Mancarella , P

    Dung, P. M, Mancarella , P. (2002). Production systems need negation as failure. IEEE Transactions on Knowledge and Data Engineering, 14(2): 336−353

    2002

  5. [13]

    Deng, XM. (2022). A New Exploration of Hegel’s Dialectics II: Negation and Reflection. Taylor and Francis

    2022

  6. [14]

    Gabbay, H

    Dov, M. Gabbay, H. W. (1999). What is Negation?. Applied Logic Series, V ol. 13, Springer. https://doi.org/10.1007/978-94-015-9309-0

    1999 doi

  7. [15]

    Fabregat, H., Duque, A., Martinez-Romo, J., et al. (2023). Negation-based transfer learning for improving 31 biomedical Named Entity Recognition and Relation Extraction. Journal of Biomedical Informatics, 138, 104279. Ferré, S. (2006). Negation, opposition, and possibility in ...

    2023

  8. [16]

    Frege, G. (1919). Negation. In: Peter T. Geach, Max Black (Eds.), Translations from the Philosophical

    1919

  9. [17]

    Gale, R. M. (1976). Negation and Non-Being. American Philosophical Quarterly Monograph, vol. 10, Basil

    1976

  10. [18]

    R., Wansing H

    Horn, L. R., Wansing H. (2020). Negation. Stanford Encyclopedia of Philosophy, Zalta, E.N. (ed.), Stanford University. http://plato.stanford.edu/entries/negation/

    2020

  11. [19]

    T., Anagnostopoulos , I., et al

    Haralabopoulos, G., Torres, M. T., Anagnostopoulos , I., et al . (2021). Text data augmentations: Permutation, antonyms and negation. Expert Systems with Applications, 177, 114769

    2021

  12. [20]

    Horn, L. R. (2004). A Natural History of Negation. Journal of Linguistics, 40(3): 426–433

    2004

  13. [21]

    Horn, L. R. (2010). The Expression of Negation. Mouton de Gruyter, Berlin

    2010

  14. [22]

    Horn, L. R. (1989). A Natural History of Negation. Chicago, University of Chicago Press

    1989

  15. [23]

    Horn, L. R. (2018). Contradiction. Stanford Encyclopedia of Philosophy, Edward N. Zalta (ed.). Stanford University. https://plato.stanford.edu/archives/win2018/entries/contradiction/ Jimé nez-Zafra, S. M., Morante, R., Martí n-Valdivia, M. T., et al. (2020). Corpora Annotated ...

    2018

  16. [24]

    Jong, D., Bagheri, A. (2022). The Case of Imperfect Negation Cues: A Two -Step Approach for Automatic Negation Scope Resolution. Lecture Notes in Computer Science 13286, Springer, 413–424

    2022

  17. [25]

    Kaneiwa, K. (2007). Negations in description logic with contraries, contradictories, and subcontraries. New Generation Computing, 25(4): 443−468

    2007

  18. [26]

    Kwiatkowski, T. (2006). On the Concept of Classical Logic. Annals of Philosophy, (2): 87−118

    2006

  19. [27]

    Kangal, K. (2019). Engels’ Intentions in Dialectics of Nature. Science & Society, 83(2): 215–243

    2019

  20. [28]

    Leszl, W. (2004). Aristotle’s Logical Works and His Conception of Logic. Topoi 23, 71–100

    2004

  21. [29]

    Concepts

    Margolis, Eric and Stephen Laurence . (2023). “Concepts”, The Stanford Encyclopedia of Philosophy (Fall 2023 Edition), Edward N. Zalta & Uri Nodelman (eds.) . Stanford University . https://plato.stanford.edu/archives/fall2023/entries/concepts/

    2023

  22. [30]

    Makkar, K. et al . (2023). Improvisation in Opinion Mining Using Negation Detection and Negation Handling Techniques: A Survey. Soft Computing: Theories and Applications, Lecture Not es in Networks and Systems 627, 799-808. https://doi.org/10.1007/978-981-19-9858-4_68

    2023 doi

  23. [31]

    A., Sohn, S., et al

    Mehrabi, S., Krishnan. A., Sohn, S., et al. (2015). DEEPEN: A negation detection system for clinical text incorporating dependency relation into NegEx. Journal of Biomedical Informatics, V ol. 54, 213-219

    2015

  24. [32]

    Mukhtar, N., Khan, M. A. (2020). Effective lexicon-based approach for Urdu sentiment analysis. Artificial Intelligence Review, 53(4): 2521–2548. https://doi.org/10.1007/s10462-019-09740-5

    2020 doi

  25. [33]

    Malinowski, G. (2007). Many-Valued Logic and its Philosophy. Handbook of the History of Logic, vol. 8: 13–94

    2007

  26. [34]

    Malpass, A., Marfori, A. M. (2017). The History of Philosophical and Formal Logic. Bloomsbury Publishing. https://doi.org/10.1023/B:TOPO.0000021384.02544.1e

    2017

  27. [35]

    Pradhan, A., Senapati , M., Sahu , P. (2021). Improving sentiment analysis with learning concepts from concept, patterns lexicons and negations. Ain Shams Engineering Journal, 13, 101559

    2021

  28. [36]

    Prllochs, N., Feuerriegel, S., Lutz , B., et al. (2020). Negation scope detection for sentiment analysis: A reinforcement learning framework for replicating human interpretations. Information Sciences, 536: 205–221

    2020

  29. [37]

    Pan, ZH. (2010). Three Kinds of Negations in Fuzzy Knowledge and Applications to Decision Making in Financial Investment. Lecture Notes in Artificial Intelligence 6422, Springer, 391−401. 32

    2010

  30. [38]

    Pan, ZH., Yang, L., Xu, J. (2011). Fuzzy Set with Three Kinds of Negations and Its Applications in Fuzzy Decision Making. Lecture Notes in Computer Science 7002, Springer, 533−542. https://doi.org/10.1007/978-3-642-23881-9_69

    2011 doi

  31. [39]

    Pan, ZH. (2013). Three Kinds of Negation of Fuzzy Knowledge and t heir Base of Logic. Lecture Notes in Artificial Intelligence 7996, Springer, 83−93. https://doi.org/10.1007/978-3-642-39482-9_10

    2013 doi

  32. [40]

    Pan, ZH. (2017). Theory and Applications of Fuzzy Sets and Fuzzy Logic with Three Kinds of Negation. Science Press, (in Chinese)

    2017

  33. [41]

    K, Paul, S

    Singh, P. K, Paul, S. (2021). Deep learning approach for negation handling in sentiment analysis. IEEE Access 9:102579–102592. https://doi.org/10.1109/ACCESS.2021.3095412

    2021

  34. [42]

    F., et al

    Slater, L.T., Bradlow, W., Motti, D. F., et al. (2021). A fast, accurate, and generalisable heuristic -based negation detection algorithm for clinical text. Computers in Biology and Medicine, 130, 104216

    2021

  35. [43]

    L., Horn, L

    Speranza, J. L., Horn, L. R. (2010). A brief history of negation. Journal of Applied Logic, 8: 277–301

    2010

  36. [44]

    Smith, Robin. (2022). Aristotle’s Logic. The Stanford Encyclopedia of Philosophy (Winter 2022 Edition), Edward N. Zalta & Uri Nodelman (eds.) https://plato.stanford.edu/archives/win2022/entries/aristotle-logic/

    2022

  37. [45]

    , Siebert, S., Stolzenburg, F

    Schon, C. , Siebert, S., Stolzenburg, F. (2021). Negation in Cognitive Reasoning. Lecture Notes in Computer Science, vol 12873. Springer, 217−232

    2021

  38. [46]

    Sethy, S. S. (2021). Fundamental Principles of Logic (The Laws of Thought). In: Introduction to Logic and Logical Discourse. Springer, Singapore. https://doi.org/10.1007/978-981-16-2689-0_5

    2021 doi

  39. [47]

    Turri, J. (2022). A Peculiar and Perpetual Tendency: An Asymmetry in Knowledge Attributions for Affirmations and Negations. Erkenntnis, 87: 1795–1808

    2022

  40. [48]

    Trillas, E. (1979). On negation functions in the theory of fuzzy sets. Stochastica: revista de matemá tica pura y aplicada, 3(1), 47-60

    1979

  41. [49]

    Wang, X., Xu, Y., He, X., et al . (2020). Reinforced Negative Sampling over Knowledge Graph for Recommendation. Proceedings of the Web Conference, WWW’20, April, 99–109. https://doi.org /10.1145/ 3366423. 3380098

    2020

  42. [50]

    Wagner, G. ( 2003). Web rules need two kinds of negation. Lecture Notes in Computer Science 2901, Springer, 33-50

    2003

  43. [51]

    Wretzel, J.I. (2014). Despair and the determinate negation of Brandom’s Hegel. Continental Philosophy Review, 47:195–216

    2014

  44. [52]

    Wybraniec-Skardowska, U. (2001). Knowledge, Vagueness and Logic. International Journal of Applied Mathematics and Computer Science, 11(3): 719−737

    2001

  45. [53]

    Wang, GJ. (1999). On the logic foundation of fuzzy reasoning. Information Sciences, (117): 47−88

    1999

  46. [54]

    Xu, D., Xu, T., Wu, S., et al. (2022). Relation-enhanced Negative Sampling for Multimodal Knowledge Graph Completion. Proceedings of the 30th ACM International Conference on Multimedia, MM’22,

    2022

  47. [55]

    October, Lisboa, Portugal. Pp. 3857–3866. https://doi.org/10.1145/3503161.3548388

  48. [56]

    Yang, C., Xiao, F. (2023) . An exponential negation of complex basic belief assignment in complex evidence theory. Information Sciences, 622: 1228–1251

    2023

  49. [57]

    Zhang, Y., Yao, Q., Chen, L. (2021). Simple and automated negative sampling for knowledge graph embedding. The International Journal on Very Large Data Bases, V ol. 30, Issue 2, 259–285

    2021

  50. [58]

    Zhu, T., Xiang, Y., Chen, Q., et al . (2023). Fine-grained biomedical knowledge negation detection via contrastive learning. Knowledge-Based Systems, V olume 272, 1100575

    2023

  51. [59]

    Zadeh, L. A. (1965). Fuzzy Sets. Information and Control, 8(3): 338−353

    1965

  52. [60]

    Zhang, LZ., Pan, ZH. (2009). Fuzzy Comprehensive Evaluation based on Measure of Medium Truth Scale. Proc. of IEEE -2009 International Conference on Artificial Intelligence and Computat ional Intelligence, 83−87

    2009

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.