Public-Data Methodology
Data-to-Rating Framework
The Data-to-Rating Framework, or D2R, is AT Worthy's methodology for transforming public digital evidence into structured, comparable, and contestable assessments.
D2R starts from a simple observation: organizations, products, public authorities, platforms, and jurisdictions now leave extensive digital traces that can reveal meaningful aspects of performance, governance, transparency, risk, and institutional capacity.
D2R does not claim that public data reveals everything. It asks when public evidence is relevant, how strong it is, how it should be interpreted, how much confidence should attach to it, and when human judgment must control the final assessment.
From public data to structured judgment
Traditional assessments often rely on questionnaires, site visits, interviews, internal documents, audits, or voluntary disclosure. These methods remain essential where privileged evidence is needed.
D2R adds another layer: public-evidence assessment. Where relevant qualities leave observable traces, D2R converts public data into qualified signals, maps them to evaluative dimensions, produces scores and ratings, and attaches evidence lineage and confidence.
Data
Public material before evaluative interpretation.
Signal
Data judged relevant to a defined construct.
Feature
An extracted property or coded element derived from signals.
Indicator
A feature or group of features mapped to an evaluative dimension.
Score
A normalized quantitative output.
Rating
A categorical interpretation of the score.
Representation
The symbol, grade, tier, badge, or label shown to users.
Confidence
The degree of evidentiary sufficiency, recency, reliability, corroboration, and interpretive stability.
Judgment
The governed human or institutional interpretation of what the output means in context.
Observable Construct Manifestation
D2R is valid only where the quality being assessed leaves public traces.
Some constructs are highly observable through public digital evidence, while others require audit, inspection, interviews, or privileged evidence. D2R works in the bounded middle, asking what can be responsibly inferred from public evidence, under what limits, and with what confidence.
Public evidence has different strength
Silence or absence may indicate no public evidence, but not necessarily poor performance.
Public assertion shows stated commitment, but weak proof of implementation.
Structured disclosure shows organized publication of relevant information.
Operational artifact shows a visible mechanism, interface, record, repository, or process.
Versioned or executable artifact can be inspected, tested, compared, or linked to an operational state.
Observed behavior over time shows consistency, responsiveness, update cadence, correction, or persistence.
Corroborated public evidence supports stronger inference because multiple sources converge on the same finding.
The D2R pipeline
Define the construct and state what is being assessed and excluded.
Set the public evidence perimeter, including admissible sources, legal boundaries, time windows, jurisdictions, languages, and exclusion rules.
Collect and version sources so evidence is attributable, timestamped, and reproducible.
Qualify signals for relevance, authenticity, recency, provenance, comparability, and resistance to manipulation.
Extract features manually, computationally, or through hybrid review.
Construct indicators and map them to dimensions with clear methodological justification.
Apply documented scoring, aggregation, weighting, threshold, cap, and non-compensability rules.
Translate scores into ratings or other user-facing outputs.
Assign confidence as part of the assessment, not as decoration.
Keep outputs contestable so stale evidence, attribution errors, missing context, or methodological misapplication can be challenged.
Confidence and Judgment
Confidence is part of the rating.
A low score and low confidence are not the same thing. A low score with high confidence may mean strong public evidence supports a negative finding. A high score with high confidence may mean positive evidence is current, relevant, and corroborated.
Any score with low confidence means the evidence may be sparse, stale, inconsistent, volatile, or weakly attributed. D2R therefore separates performance judgment from evidence sufficiency.
Human Authority
Automation does not replace judgment.
D2R can use automation and AI to discover sources, classify documents, extract signals, detect changes, compare evidence, summarize public materials, and flag anomalies.
Human control remains required for construct design, evidence admissibility, scoring rules, thresholds, methodology changes, disputed findings, high-stakes use, and final interpretation.
Methodological Layer
Relationship with Digital Worthiness and AI Worthiness
D2R is the methodological layer behind AT Worthy's rating work. Digital Worthiness and AI Worthiness define what is being assessed. D2R defines how public evidence can support the assessment, strengthening AT Worthy's ability to produce scalable, defensible, updateable, and contestable ratings.
Where D2R can be applied
Governance safeguards
The Promise of Data-to-Rating
From public evidence to defensible structured judgment.
Assessment is moving from episodic review toward continuous evidence awareness. D2R allows ratings to become more dynamic, evidence-linked, and transparent while identifying what public evidence shows, what it does not show, and when deeper review is needed.
AT Worthy uses the Data-to-Rating Framework to strengthen the credibility, transparency, and scalability of its assessment work across Digital Worthiness, AI Worthiness, and future frameworks.
Measure Your Worthiness
Individuals, organizations, and institutions increasingly depend on digital and AI systems to operate, deliver services, and make consequential decisions. AT Worthy provides independent evaluation, trusted ratings, and AI-driven analysis to assess how these systems perform, how they can be trusted, and where they require improvement.
Get Started
AT Worthy is a sponsor of the UNICODE as a Lifelong and Unique Gold Adopter of the character « @ » the Digital Rating's symbol.
AT Worthy proudly stands as a founding member of the GliaNet Alliance, joining a coalition committed to ethical technology and digital trust.