Keywords — Controlled vocabulary

Learn how keywords from the footage.one taxonomy work and how to use them effectively.

Overview

Keywords are the central tool for making media findable in footage.one. Unlike simple text terms, keywords come from a controlled taxonomy — a structured vocabulary with translations, synonyms and relationships between concepts.

Keywords and Labels complement each other: keywords provide precise, normalised descriptions; labels are flexible tags like hashtags.

What makes keywords special

Multilingual

Every keyword is stored in multiple languages. If you assign a keyword in German, it is automatically displayed in the correct translation for English-speaking users — and vice versa.

Synonyms and spelling variants

The taxonomy knows alternative names:

  • "Muhammad Ali" can also be found via "Cassius Clay"
  • "Gorbachev" and "Gorbatschow" lead to the same keyword
  • "Auto", "PKW" and "Personenkraftwagen" all point to the same concept

Relationships

Keywords are related to one another. The taxonomy recognises, among others:

  • Part of / Has parts — e.g. "Engine" is part of "Car"
  • Instance of — e.g. "Eiffel Tower" is an instance of "Tower"
  • Subclass of — e.g. "Sports car" is a subclass of "Car"
  • Geography — e.g. "Paris" belongs to "France"
  • Related — general thematic proximity

These relationships automatically broaden search results: searching for "France" can also find assets tagged with "Paris".

Assigning keywords

Manually

  1. Open an asset and switch to the tagging area
  2. Start typing in the search field — suggestions appear after the third character
  3. Select the appropriate keyword from the list
  4. The keyword is assigned immediately

Suggestions show not only the keyword name but also a description and alternative labels, so you can select the right term.

Search modes

Several modes are available when searching for keywords:

  • Exact — finds only exact matches
  • Starts with — finds keywords beginning with the search term
  • Contains — finds keywords containing the search term
  • Deep search — also searches descriptions and synonyms

AI-assisted

footage.one offers automatic tagging powered by AI:

  • Object recognition — detects objects, people, scenes and locations in images
  • Visual similarity — adopts keywords from visually similar media
  • Colour recognition — detects dominant colours
  • Location keywords — derives keywords from geo-metadata
  • Description analysis — extracts keywords from description texts

You can review the AI suggestions and accept or reject them with a single click.

Note: AI auto-tagging is a subscription feature and is not available in all plans.

Keywords at different levels

Keywords can be assigned at different levels:

Level Description
Asset Keywords apply to exactly this media item
Episode Keywords apply to all assets in the episode — ideal for shared descriptions
Album Keywords for a thematic collection
Time range For videos, keywords can be applied to specific time ranges

Assets inherit keywords from their episode and albums. This means: a keyword assigned at episode level makes all assets in that episode findable via that keyword — without having to assign it individually.

Excluding keywords

If a suggested keyword does not fit, you can exclude it:

  1. Click the exclusion icon next to the keyword
  2. Choose the type of exclusion:
    • Single keyword — only this keyword is excluded
    • Entire class — all keywords of this category are excluded (e.g. all instances of "Tower")

Excluded keywords no longer appear in suggestions.

Free-text keywords (Freewords)

If you enter a term that does not exist in the taxonomy, it is saved as a Freeword. Freewords work like regular keywords but are not normalised — they have no translations or relationships.

Administrators can subsequently map Freewords to a taxonomy keyword using the Mapping Tool. This ensures consistent tagging throughout the entire archive.

Tips

  • Start at episode level — assign common keywords at episode level first, then refine individual assets
  • Think associatively — not just describe what is visible, but also what might be associated
  • Use AI as a starting point — let AI make suggestions and add manually
  • Use mappings — regularly map Freewords to taxonomy keywords