Data Tag Homogenization#

One of the objectives of the Data Curator is to homogenize the names of the concepts represented by different data provider tags. Below are the prefixes we assign to each “block” or category of data rows, followed by the rules used to construct the final tag names.

Prefix Mapping#

  • fbs_ – Fundamental Balance-Sheet data rows

  • fcf_ – Fundamental Cash-Flow data rows

  • fis_ – Fundamental Income Statement data rows

  • f_ – Fundamental Common data rows (e.g., identifiers shared across statements)

  • m_ – Market data (daily, including unadjusted, split-adjusted, and dividend-and-split-adjusted)

  • d_ – Dividend data rows

  • s_ – Split data rows

Tag Construction Rules#

  1. Single Snake_Case Identifier All tags must be a single “word” in snake_case (no spaces or hyphens), so they can be used directly as Python parameter or variable names.

  2. Full Concept Spelled Out The tag spells out the full concept in underscore-joined words. - Avoid acronyms and abbreviations unless they are industry standards. - Examples:

    • fis_revenue (not fis_rev)

    • fbs_shareholders_equity (not fbs_sq_eq)

  3. Natural Word Order Words appear in the order they would naturally be spoken: modifiers (adjectives) come before the main noun. - Example:

    • nonoperating_income (instead of income_nonoperating)

  4. Hyphenated Prefixes Treat any hyphenated phrase as a non-hyphenated sequence.

    • Example: - Original label: “Non-Operating Income” - Tag: nonoperating_income

  5. Avoiding Ambiguity with Opposing Subconcepts

    If a concept can represent two opposing subconcepts depending on sign (e.g., “Income (Loss)”), only the primary subconcept is used. - Example:

    • Use fis_income rather than fis_income_loss.

  6. Restricted Use of “net” Use “net” only when the subtraction is unambiguous (i.e., net after subtracting a directly opposing concept).

    • Examples: - fis_net_proceeds implies proceeds minus direct costs. - fis_net_cash_from_operating_activities implies operating cash inflows minus operating cash outflows.

    If “net” would be ambiguous, spell out the offset using _after_ when practical.

    • Example: use after_tax instead of net_of_tax.

  7. Reordering to Remove “of” Remove the conjunction “of” unless it is part of a widely recognized concept name. - Example:

    • Keep cost_of_revenue because “Cost of Revenue” is standard.

    • Convert “Depreciation of Assets” to asset_depreciation.

  8. Aligning with US GAAP Taxonomy When a tag matches a US GAAP taxonomy concept, name it as closely as possible while enforcing the above rules. If the official taxonomy label omits a key word, include it.

    • Example: - Taxonomy concept “Interest Income (Expense), Operating” - Tag: fi_interest_income_operating (rather than the shorter fis_interest_income_expense).

  9. Singular vs. Plural Use the singular form for a specific subtype (e.g., fis_interest_expense). Use the plural form for general aggregates (e.g., fis_operating_expenses).

  10. Preferred Terminology Use stockholder instead of “shareholder.” Use noncontrolling_interest instead of “minority_interest.” Only use total when the concept is explicitly the sum of multiple subcomponents (e.g., fbs_total_debt_including_capital_lease_obligations).

  11. Disambiguating Cash-Flow Terminology For cash-flow concepts involving net “Increase/Decrease,” use the term change: - Example: fcf_inventory_change (rather than “Inventory Increase/Decrease”).

    For cash inflows, use proceeds; for outflows, use payments:

    • Examples: - fcf_common_stock_issuance_proceeds - fcf_preferred_stock_dividend_payments

By following these rules and using the prefixes listed above, Data Curator achieves a consistent, unambiguous naming convention across all data-provider tags.