ROS / Nea Polis

BIS & NIS

Specialized analysis with explicit boundaries: business system and current-affairs module.

Coupled system / 0.3.1

Nea Polis BIS

What does a company do, how is it organized, what does it offer to whom, and which economic or operational relationships are evidenced?

Separate Business Intelligence System coupled to ROS, at v0.3.1. OrganizationArchetype and ApplicabilityProfile determine which semantics for orders, territories, markets, assets and metrics apply. The Common Meta-Kernel provides shared rules; ROS 1.7.2 is not a BIS version upgrade.

  • Facts, Measures, Metrics and KPIs have their own aggregation and temporal semantics.
  • Portfolio, customers, competition and public perception complement the quantitative layer.
  • Personal roles, legal lineage and operational relationships do not automatically imply ownership relationships.
  • Recursion ends at explicit scope boundaries; missing commercial terms remain unknown.
  • NOT_APPLICABLE and UNKNOWN remain distinct; a metric is not an epistemic claim.
  • Content completeness and actually executed graph closure are different states.

Documentation basis: BIS v0.2 -> v0.3 (2026-08-29). Source type and specific verification limits are recorded in the provenance register; citing a source does not independently validate every claim.

Project case study

ROS domain module / Migration 1.0 · SARM 1.1 / RC1

ROS.NIS / News Intelligence

How do different information environments respond to an event over time, and which divergences warrant further investigation?

ROS.NIS remains a domain module within the shared research state. Migration 1.0, SARM 1.1 / RC1 and Speculative Layer/Narrative Delta form the documented composite RC1 profile. The current global resolver incorporates the applicable ROS core 1.7.2 without promoting NIS itself. A new integrated native schema/production release, full-corpus cutover and long-term calibration remain unconfirmed.

  • A signal or index is an analytical intermediate product, not a truth or a confirmed prediction.
  • Original measurements and raw data are retained alongside derived series.
  • Snapshots are append-only; corrections reference predecessors rather than overwriting past knowledge.
  • Platform samples are not representative political population groups.
  • Coverage, baseline, time window, method version and uncertainty belong in the dataset.
  • Measurement Confidence, Integrity Confidence and Manipulation Risk are not collapsed into one value.
  • Speculative claims remain separate from the situational assessment until evidence-based resolution; promotion serialization is still unresolved.

X / Reddit

Separate discourse samples: frames, salience, sentiment and momentum. Political classification is a sampling assumption that requires verification.

Prediction Markets

Market question, resolution conditions, probability, liquidity and shifts in capital-weighted expectations.

Contrarian

Divergent positioning and information anomalies with alternative explanations; not evidence of insider knowledge.

Public Representation Index

Role expectations, observed presence, assumption of responsibility and reception; not a popularity index.

SARM Integrity & Observability

Discourse Structure, Administrative Motion, Bureaucratic Friction, Omission Residual and multivariate anomalies as separate experimental sensors; no automatic attribution.

Speculative Layer / Narrative Delta

Approved working state for open claims and comparison of Official Narrative, Independent Evidence and Speculative Layer. Status, provenance, propagation, contradiction and testable consequences remain historically traceable.

Documentation basis: NIS Latest · Unchanged RC1 Profile with ROS Core 1.7.2 (2026-10-15). Source type and specific verification limits are recorded in the provenance register; citing a source does not independently validate every claim.

Project case study