Genre taxonomies
A genre taxonomy organizes classification concepts and their labels or relationships, allowing consistent membership beyond matching display strings.
Genre labels help readers browse works with related styles or traditions. A taxonomy can organize those classification concepts, their labels and any defined broader/narrower relationships.
A concept can have several labels
An original vocabulary model is:
| Concept key | English label | French label |
|---|---|---|
| C1 | Drama | Drame |
| C2 | Mystery | Mystère |
C1's labels name one classification concept. Counting Drama and Drame as unrelated categories because the strings differ can split the same intended membership.
A vocabulary can also define preferred and alternate labels. Deciding that Sci-Fi and Science Fiction are synonyms requires that vocabulary relationship, rather than a universal rule to merge every similar string.
Membership can overlap
A mystery drama can belong to both C1 and C2. That does not require two copies of the movie.
A browse view that counts each genre independently can therefore have category totals greater than the number of distinct works. Faceted browsing develops the corresponding set arithmetic.
In Emby, compare imported genre entries with the selected browse/filter result. A display list can be localized or curated while the intended classification remains stable.
Genre, subject and rating answer different questions
Railway can be a useful subject keyword for a story without being a universal film genre. A parental classification describes another relationship; it does not automatically replace Drama or Mystery.
A recommendation system can use several such features, but their meanings should remain explicit. Combining them into one label string makes later filtering harder to interpret.
Vocabularies can change
A provider, an editorial catalogue and a user's personal organization can use different category granularity. A broader concept and a more specific subgenre need not be contradictions.
SKOS supplies a reference model for concepts and labels; it does not mandate that every media library implement the full model. The relevant decisions are which vocabulary the entries belong to, how aliases/localized labels are mapped and what membership a query tests.