Problem thread

Ontology against folksonomy

Should shared terms be designed by experts in advance, or emerge from the words people actually use?

issue
vocabulary
status
open
steps
11, 1876–2018
threads
all problem threads

Libraries organized knowledge with expert-designed classifications, and knowledge engineers carried the idea into formal ontologies. Critics argued that metadata is unreliable and that hierarchies are a habit of the physical shelf, while tagging sites let users label things in their own words. Schema.org and Wikidata settled into a middle course: shared vocabularies that a community extends through an open process as needs appear. Large communities found a middle way: free tags curated by volunteers, as on OpenStreetMap and Archive of Our Own.

Lineage

  1. 1876
    Dewey Decimal Classification

    Sorts all knowledge into ten main classes with decimal subdivisions, fixed by experts in advance. [src]

  2. 1933
    Colon Classification

    Builds a subject's number from facets joined by colons, rather than placing it in a single fixed hierarchy. [src]

  3. Jun 1993
    A Translation Approach to Portable Ontology Specifications

    Defines an ontology as 'an explicit specification of a conceptualization', to be shared between knowledge systems. [src]

  4. 26 Aug 2001
    poses writing ★ seminal
    Metacrap: Putting the torch to seven straw-men of the meta-utopia

    Reliable metadata is unattainable: people lie, are lazy and stupid, cannot describe themselves; schemas are not neutral, metrics influence results, and there is more than one way to describe something. [src]

  5. Jul 2004
    proposes concept
    Folksonomy

    Thomas Vander Wal coins the word in a July 2004 discussion of services such as Flickr and del.icio.us, where users tag things in their own words and categories emerge from the tags. [src1][src2]

  6. 9 Aug 2004
    OpenStreetMap

    OpenStreetMap lets mappers use any key=value tags they like, steered by a recommended scheme on a wiki: a folksonomy with conventions, at the scale of a world map. [src]

  7. 2005
    Ontology is Overrated: Categories, Links, and Tags

    Hierarchical classification optimizes for books on a shelf, as the Dewey treatment of religions shows; on the Web 'there is no shelf', and links and free tagging work better. [src]

  8. 14 Nov 2009
    workaround project
    Archive of Our Own

    Archive of Our Own lets writers tag freely and has volunteer tag wranglers link synonymous tags to canonical ones, keeping the writers' words while adding structure. [src1][src2]

  9. 2 Jun 2011
    Schema.org

    A shared vocabulary started by the major search engines. Anyone can propose changes through public lists and GitHub; a community group reviews them and a steering group approves releases. [src]

  10. 29 Oct 2012
    Wikidata

    The vocabulary grows by community agreement: before a new property is created it has to be proposed and discussed, then a property creator or administrator adds it. [src]

  11. 2018
    measures writing
    Curated Folksonomies: Three Implementations of Structure through Human Judgment

    Julia Bullard compares collections that structure their tags through human judgment and names the approach the curated folksonomy, a middle way the debate had not named. [src]

Open now

Under each question, the answers people have put forward or built. The catalogue gives no odds on them.

  • Can a vocabulary that grows from use stay consistent enough for machines to rely on? [src1][src2]
    Schema.org and Wikidata both add terms through open community processes rather than a fixed design.
    • Terms grow from use, then shapes check them. Wikidata
      Wikidata adds properties through community proposals and describes the expected shape of items in EntitySchemas written in ShEx.[src1][src2]
    • Community proposals, with a steering group deciding. Schema.org
      Schema.org takes proposals from its community group, and its steering group decides what enters the core.[src]

How to contribute