Semantic gaps and open research

The central hypothesis is a need for more explicit scientific justification of cross-scale and cross-support transformations. This is a project research area, not a claim that no ontology or literature already addresses it.

Describe a thing versus justify a transformation

Existing resources offer substantial capabilities for entities, properties, observations, environments, geography, time, units and provenance. The difficult question is how to record why a specific scientific transformation is justified:

individual leaf measurements
  → species trait distribution
  → community trait distribution
  → plot-scale property
  → 30 m remotely sensed observation
  → landscape representation

Each arrow can introduce sampling, membership, statistical, physical or modeling assumptions. A list of entity labels and part-of relations does not supply those assumptions. A plot and a pixel can overlap while observing different quantities with different response functions.

What must be compared

For each transformation, inspect source/target entity definitions, population or community membership, methods, sampling design, spatial/temporal weighting, completeness, uncertainty, transfer assumptions and validation evidence. Distinguish a proposed derivation from observed data. Record the exact reviewed recipe and human decision where required.

PCO, GeoSPARQL, OWL-Time, QUDT and PROV-O can each contribute useful descriptions. None alone proves a cross-support derivation scientifically justified. The inventory's limitations are provisional scope assessments, not declarations that those resources cannot be extended or composed.

Continued literature review

Investigate existing observation-support models, statistical aggregation semantics, remote-sensing response models, ecological sampling models, scientific workflow provenance and evidence/argumentation frameworks. Compare actual axioms and worked examples before proposing a project concept. The Bari Manifesto and OBOE are relevant precedents, not an exhaustive review.

The benchmark's semantic_gaps.parquet counts assessed inventory coverage and records this hypothesis as not measured. It cannot establish the absence of an external ontology or rank the scientific validity of transformations.

Conservative new-concept decision process · Proposed semantic profile