taxonomy.md

Taxonomy Guide

The Tapiro Taxonomy is a hierarchical classification system used to categorize products, services, and user interests. Accurate use of this taxonomy is essential for the effectiveness of Tapiro's AI models in generating meaningful user preferences.

Purpose

• Standardization: Ensures that data submitted by different stores uses a consistent language for products and interests.
• AI Training: Provides structured input for the machine learning models that analyze user behavior and build preference profiles.
• Preference Granularity: Allows for preferences to be understood at different levels (e.g., general interest in "Electronics" vs. specific interest in "Smartphones" with a preference for "Apple" brand).

Structure

The taxonomy consists of:

• Categories: Broad groupings (e.g., "Electronics", "Fashion", "Home").
• Sub-categories: More specific groupings within a parent category (e.g., "Smartphones" under "Electronics"). Categories have unique IDs (e.g., "101") and names (e.g., "Smartphones").
• Attributes: Characteristics relevant to a specific category (e.g., "brand", "color", "size" for "Smartphones"). Attributes have defined possible values.

Example Snippet (Conceptual):

version: "1.0.0"
categories:
  - id: "100"
    name: "Electronics"
    # ... attributes for Electronics ...
  - id: "101"
    name: "Smartphones"
    parent_id: "100"
    attributes:
      - name: "brand"
        values: [Apple, Samsung, Google, ...]
      - name: "color"
        values: [black, white, blue, ...]
      # ... other smartphone attributes ...
  - id: "200"
    name: "Fashion"
    # ... attributes for Fashion ...
  - id: "201"
    name: "Clothing"
    parent_id: "200"
    attributes:
      - name: "type"
        values: [shirts, pants, dresses, ...]
      - name: "material"
        values: [cotton, polyester, wool, ...]
      # ... other clothing attributes ...