Strategic Categorical System (SCS): Faceted Navigation with Polynyms

Overview
The Strategic Categorical System (SCS) organizes entities by determining whether their relationship to a headword is a type, part, or step. Body facets (the individual related entities) are grouped into polynyms, which may consist of one facet (mononym), two (bionym), three (trionym), or more. The system uses computational tools such as TF-IDF, cosine similarity, typo metrics, and phonetic similarity to classify relationships reliably and consistently.

1. Headword and Body Facets
The headword is the canonical entity name (e.g., a normalized outlet name). Body facets are the variants, related items, or associated forms linked to the headword. The SCS determines how each body facet relates to the headword: as a type, part, or step.

2. Type Classification
A facet is classified as a type when it represents a different kind or category under the same headword.
Conceptual meaning: species under a genus.
Indicators:

  • Distinctive TF-IDF features that differ strongly from the headword.
  • Low cosine similarity (different conceptual direction).
  • Phonetic similarity is not relevant.
  • Large edit distance.
    Usage example: different kinds of outlets under the umbrella concept “Media Outlet.”

3. Part Classification
A facet is classified as a part when it represents a functional component of the headword.
Conceptual meaning: mereological decomposition (constituent elements).
Indicators:

  • High TF-IDF similarity on context terms.
  • Moderate cosine similarity.
  • Edit distance may vary significantly.
  • Non-name metadata (location, URL domain, organizational structure) strongly informs classification.
    Usage example: departments, desks, or structural divisions of an outlet.

4. Step Classification
A facet is classified as a step when it represents a successive step, version, or form of the headword.
Conceptual meaning: diachronic identity or developmental stages.
Indicators:

  • Very high TF-IDF overlap.
  • Very high cosine similarity (same conceptual direction).
  • Low edit distance for stems.
  • Strong phonetic similarity.
  • Domain or URL identity reinforces classification.
    Usage example: platform versions (web, app, broadcast) or rebranded forms of the same outlet.

5. Polynyms and N-onyms
Body facets connected to the same headword form a polynym. The count of connected facets determines whether the polynym is a mononym (1), bionym (2), trionym (3), tetranym (4), etc. The number is descriptive only; classification depends on the type/part/sept relationship.

6. Computational Tools and Their Role in SCS

TF-IDF: Detects distinctive terms that separate types from each other (differentia).
Cosine similarity: Measures directional alignment in semantic space; distinguishes types from parts and septs.
Phonetic similarity: Establishes identity continuity for name variations; useful for septs.
Typo metrics (e.g., Levenshtein): Measures the transformation distance between appearances; assists in cleaning and unifying facets before classification.

7. Summary of Conceptual Logic

Type: difference of essence.
Part: difference of function.
Step: difference of form.

Computational tools provide quantitative signals that support these philosophical distinctions. The SCS supplies the conceptual rules; the computational system supplies the measurements.

Assessment of the Current Tools

The tools being used (TF-IDF, cosine similarity, phonetic algorithms, and typo-distance metrics) are not sufficient to fully implement the Strategic Categorical System (SCS). However, they are the correct foundational starting point.