Documentation

How to Build a Simple, Usable Photo Keyword Thesaurus

Build a controlled vocabulary without becoming an archivist, and without drowning your photo library in inconsistent keywords.

Metadata

The problem with free keywords

At first, you type keywords as you go. Then, six months later, the same idea exists in five variants: portrait, portraits, Portrait, model, person. Search becomes a guessing game: which one did I use that day? Free keywords are fast to add and slow to exploit. A controlled vocabulary, a thesaurus, fixes this without requiring you to become a librarian.

Free tag vs. controlled term

A free tag is text typed on the fly: nothing guarantees its consistency from one time to the next. A controlled term belongs to a reference list: you choose it rather than retype it. The difference is decisive at search time: with a controlled vocabulary, searching a term retrieves every relevant image, not just the ones where you typed the right variant.

The building blocks of a simple thesaurus

No need for a heavy machine. Three mechanisms are enough:

  • Synonyms: sea, ocean, seaside all point to the same term. Type however you like; search finds everything.
  • Broader and narrower terms: a light hierarchy, from general to specific, animalbirdgull. Searching the broad term surfaces the specific ones.
  • Categories: separating places, people, techniques, and themes avoids dumping everything into one bag of words.

Leaning on a method: QQOQCP

To avoid missing dimensions, the old journalistic method QQOQCP (Who, What, Where, When, How, Why) offers a simple grid. Each question suggests a family of terms: who (people), what (subjects, objects), where (places), when (period, season), how (technique, process), why (context, intent). You get a balanced vocabulary without reinventing documentary classification.

Adopt it progressively

Don’t build a complete thesaurus before tagging a single image. Do the reverse: start tagging, then promote recurring terms to controlled status as you go, attaching their variants as synonyms. The vocabulary grows with use, instead of being a theory imposed from above.

Common mistakes

  • Too many terms: an 800-entry thesaurus is no longer usable. Stay at the level of granularity you actually exploit.
  • Mixing singular and plural, cases: set one simple rule and hold to it.
  • Confusing theme and description: the thesaurus classifies, it doesn’t narrate, narration goes in the caption.

In practice

  • Decide on one reference term per idea, and attach synonyms to it.
  • Organize into a few categories (places, people, techniques, themes).
  • Adopt the vocabulary progressively, promoting recurring tags.
  • Keep it small and alive rather than big and dead.

Obscura Flow combines free tags, a thesaurus (synonyms, broader/narrower terms), and QQOQCP indexing: you type freely, then consolidate, without ever giving up the simplicity you started with.