Boundary Transparency Index (BTI) v1.0 Supplementary Note: F5 Uncertainty as a Fuzzy Linguistic Extension Pathway
Year: 2026
Type: article
Abstract: Correction notice, 7 July 2026 This record should be treated as an archived methodological note. A later source-chain reconciliation of the ATRS-HS46 strict scoring table showed that the record-level fuzzy F5 example set used in this note is not a valid F5=1 set. A formal correction note has been published as: BTI F5 Fuzzy Extension Correction Note: Source-Chain Reconciliation of ATRS-HS46 Uncertainty Scoring, v0.2.DOI: 10.5281/zenodo.21237907 The canonical ATRS-HS46 strict F5 distribution is: F5 = 29 score 0; 17 score 1; 0 score 2; mean 0.37. The earlier three-record fuzzy example is superseded: TrustID and DHSC/NHS Digital AI Writing Assistant score F5=0 under the strict event-level rule; only Whitemail scores F5=1. Therefore, FUZZY-001 and FUZZY-002 should not be cited as empirical findings. The core ATRS-HS46 findings are unchanged: N=46, mean strict PLAN/BTI score 6.87/12, median 7, range 5–9, and F5 absent in 29/46 records. The surviving conceptual insight is revised as follows: withheld uncertainty metrics should be analysed as a subcondition within F5=0, not as a fuzzy refinement within F5=1. This record remains useful only as a provenance record for the methodological development of the BTI F5 fuzzy-extension idea. It should not be cited as evidence for the superseded three-record fuzzy F5 analysis, the TrustID F5=(s1,+0.25) claim, or the claim that the analysed three records constituted the F5=1 set. This supplementary technical note documents a post-validation methodological extension pathway for the Boundary Transparency Index (BTI), focusing on F5 Uncertainty as a candidate fuzzy linguistic extension. BTI remains a six-field ordinal 0-12 reconstructability instrument for evaluating whether decision records preserve the minimum informational content required for post-event accountability, auditability, and contestability. This note does not modify the current BTI scoring scheme and does not introduce a newly validated fuzzy instrument. The purpose of the note is to preserve a bounded methodological insight emerging from IJCIS field-positioning work: among the six BTI dimensions, F5 Uncertainty is the most natural candidate for future fuzzy-system extension because decision-point uncertainty is graded, linguistic, and context-sensitive rather than purely binary or crisp. The validation study motivating this note found that F5 produced the largest DER advantage while also showing the widest rater-calibration difference, suggesting that crisp ordinal scoring is useful but may be less expressive for this dimension than for more document-presence-oriented fields. The proposed extension pathway treats decision-point uncertainty as a possible fuzzy linguistic variable, with future work potentially drawing on two-tuple linguistic representation models, fuzzy aggregation operators, or related fuzzy decision-making methods. The note explicitly limits its claim: no membership functions are fixed, no fuzzy rescoring is reported, and no fuzzy validation study is claimed. This record is intended as a supplementary methodological anchor for future BTI development, especially for computational intelligence, fuzzy systems, multi-criteria decision-making, and decision-support-system research contexts.
Author: So, Hon Bor
Language: English
Subfield: Statistics, Probability and Uncertainty
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7162525630