Lady’s Gitweb - Blog/blob - 2023-05-26/federated_tagging/#entry.rdf

   1 <awol:Entry
   2         xml:lang="en"
   3         xmlns:awol="http://bblfish.net/work/atom-owl/2006-06-06/"
   4         xmlns:dc11="http://purl.org/dc/elements/1.1/"
   5         xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
   6         xmlns:sioc="http://rdfs.org/sioc/ns#"
   7 >
   8         <dc11:title>Federated Tagging of Fanworks</dc11:title>
   9         <dc11:date>2023-05-26T17:27:26-07:00</dc11:date>
  10         <dc11:abstract rdf:parseType="Markdown"><![CDATA[
  11 Fans like to use <cite>A·O·3</cite>’s tagging system as an excuse for
  12 why decentralization is impossible. But actually, there’s no reason why
  13 having a centralized tag authority should require centralization in
  14 fannish repositories as well.
  15 ]]></dc11:abstract>
  16         <sioc:content rdf:parseType="Markdown"><![CDATA[
  17 As [recent conversations regarding a·i][satsuma:25796] and [racism in
  18 fandom][end_otw_racism] continue to exert pressure on the idea that the
  19 [Organization for Transformative Works][OTW] is in any way a good
  20 steward of the 11 million fanworks on their website [<cite>Archive of
  21 Our Own</cite>][AO3], some fans are again beginning to wonder if a
  22 centralized repository controlled by a single organization is truly the
  23 best solution for a community as diverse as fandom.
  24
  25 There are a few reasons why one might say it is, but in this post I
  26 want to confront one: tagging. Tags (short labels which have been
  27 “wrangled” and organized into a hierarchial concept scheme) are the
  28 life·blood of search, filtering, and discovery on <cite>A·O·3</cite>,
  29 and powering this system is a huge cadre of volunteers whose duty it is
  30 to ensure that each label an author assigns to their work is organized
  31 correctly. <cite>A·O·3</cite>’s tagging system isn’t perfect—far from
  32 it—but the fact remains that a small, independent fan repository is
  33 unlikely to ever be able to muster the level of volunteer effort needed
  34 to build a tagging system of comparable utility. So, the argument goes,
  35 posting works to <cite>A·O·3</cite> is, at the very least, a necessary
  36 evil.
  37
  38 I think this argument suffers from a failure of imagination. Even if we
  39 agree on the premise that tag wrangling work is labour‐intensive and
  40 fannish needs are best accommodated by a large, centralized tag
  41 authority, there is absolutely no reason why that tag authority
  42 couldn’t be shared by multiple small, independent repositories instead
  43 of being entirely controlled by one big one. We just need a way to get
  44 the tags from point A to point B.
  45
  46 This blogpost describes one such way. It uses [the <cite>Tagging</cite>
  47 vocabulary][Tagging] to describe and federate tags from one centralized
  48 authority—which could be a large, volunteer‐run tag‐wrangling service
  49 in the model of <cite>A·O·3</cite>—out to potentially infinite
  50 independent repositories. It’s not necessarily a complete solution, and
  51 it’s unlikely that a system built with these technologies would work
  52 exactly like things do at <cite>A·O·3</cite>—but this is not
  53 necessarily a bad thing.
  54
  55 ## Tag Model
  56
  57 The <cite>Tagging</cite> vocabulary is an R·D·F ontology, which is to
  58 say it is a collection of terms and definitions which might be used to
  59 convey information between computers on the Web. It’s not a syntax or
  60 a protocol, although the conventions of the present day dictate that
  61 this information should probably be conveyed in [J·son‐L·D][JSON-LD]
  62 via normal H·T·T·P·S `GET` requests. A tag might look as follows :—
  63
  64 ```json
  65 { "@context": "https://ns.1024.gdn/Tagging/context.jsonld"
  66 , "@id": "https://ladys.example/tag:ladys.example,2023-05-23:5NJ-8SKG"
  67 , "@type": "SexualRelationshipTag"
  68 , "prefLabel":
  69   { "literalForm":
  70     { "@value": "Shadow × O·C"
  71     , "@language": "en" } }
  72 , "altLabel":
  73   { "literalForm":
  74     { "@value": "Sexual Relationship: Shadow the Hedgehog & Original Character"
  75     , "@language": "en" } }
  76 , "involves":
  77   [ "https://ladys.example/tag:ladys.example,2023-05-23:N54-M773"
  78   , "https://ladys.example/tag:ladys.example,2023-05-23:V2G-36D8" ] }
  79 ```
  80
  81 This object describes the tag
  82 `tag:ladys.example,2023-05-23:5NJ-8SKG`—or, as humans might refer to
  83 it, ‹ Shadow × O·C ›. It is a sexual relationship tag involving two
  84 other tags (`tag:ladys.example,2023-05-23:N54-M773` and
  85 `tag:ladys.example,2023-05-23:V2G-36D8`—presumably ‹ Shadow › and
  86 ‹ Original Character ›), and it has both a preferred label and an
  87 alternate one provided. If `ladys.example` were a real tag authority
  88 and this were a real tag, any repository would be able to fetch this
  89 U·R·L and retrieve the tag, its label, and its relationships.
  90
  91 If <cite>A·O·3</cite> provided this kind of service, we could begin
  92 building fanfiction repositories which leveraged their tagging system
  93 *today*.
  94
  95 ## Works and Tag Labels
  96
  97 Of course, simply having dereferencable tags on the internet is only
  98 half of the struggle. We also need a means of associating them with
  99 works. In the <cite>Tagging</cite> vocabulary, this happens through a
 100 layer of indirection, because the labels that authors provide often do
 101 not match up exactly with the tag authority’s preferred labels for the
 102 tag. The way this works is as follows :—
 103
 104 - The work itself *has a tag label collection*, probably an ordered
 105   one.
 106
 107 - This tag label collection then *contains tag labels*, which provide
 108   the author‐assigned labels attached to the work.
 109
 110 - Labels may then *label* a specific tag, linking to the formal
 111   definition of the tag in some authority.
 112
 113 How this all gets actually managed in practice is up to the repository
 114 in question; it’s certainly possible to use tags with a simpler (or
 115 more complex) model than this, and questions of how works *themselves*
 116 might get federated is a bit out·of·scope for this discussion. However,
 117 my own immediate target is building an [Atom feed][RFC4287] of
 118 fanfiction I’ve written; in this case, the tags might be added using
 119 [AtomTriples][draft-nottingham-atomtriples-00] like so :—
 120
 121 ```xml
 122 <feed
 123   xmlns="http://www.w3.org/2005/Atom"
 124   xmlns:at="http://purl.org/syndication/atomtriples/1"
 125   xmlns:html="http://www.w3.org/1999/xhtml"
 126   xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
 127   xmlns:t="https://ns.1024.gdn/Tagging/#"
 128   xmlns:xl="http://www.w3.org/2008/05/skos-xl#"
 129   xml:lang="en"
 130 >
 131   <title>Lady’s Fanfiction</title>
 132   <link rel="alternate" type="text/html" href="https://fanfiction.ladys.example/"/>
 133   <updated>2023-05-23T12:53:37-07:00</updated>
 134   <author>
 135     <name>Lady</name>
 136   </author>
 137   <id>https://fanfiction.ladys.example/tag:fanfiction.ladys.example,2023-05-23:</id>
 138   <entry>
 139     <title>fucky wucky shadow daddy dom me</title>
 140     <link rel="alternate" type="text/html" href="https://fanfiction.ladys.example/tag:fanfiction.ladys.example,2023-05-23:EQW-D8HX"/>
 141     <id>https://fanfiction.ladys.example/tag:fanfiction.ladys.example,2023-05-23:EQW-D8HX</id>
 142     <updated>2023-05-23T12:53:37-07:00</updated>
 143     <summary type="text">
 144       big bad shadow the hedgehog thinks hes the ultimate lifeform. but
 145       hes actually kind of short. can he still dom a hot dyke like me?
 146     </summary>
 147     <at:md>
 148       <!-- The tag label collection for this entry -->
 149       <t:hasTagLabelCollection rdf:parseType="Resource">
 150         <t:hasTagLabelList rdf:parseType="Collection">
 151           <t:TagLabel xl:literalForm="fuuck me shaddw">
 152             <t:labels>
 153               <!-- Shadow × Original Character -->
 154               <rdf:Description rdf:resource="https://ladys.example/tag:ladys.example,2023-05-23:5NJ-8SKG"/>
 155             </t:labels>
 156           </t:TagLabel>
 157           <t:TagLabel xl:literalForm="daddy th ehedgeog">
 158             <t:labels>
 159               <!-- Shadow -->
 160               <rdf:Description rdf:resource="https://ladys.example/tag:ladys.example,2023-05-23:N54-M773"/>
 161             </t:labels>
 162           </t:TagLabel>
 163           <t:TagLabel xl:literalForm="me">
 164             <t:labels>
 165               <!-- Original Character -->
 166               <rdf:Description rdf:resource="https://ladys.example/tag:ladys.example,2023-05-23:V2G-36D8"/>
 167             </t:labels>
 168           </t:TagLabel>
 169         </t:hasTagLabelList>
 170       </t:hasTagLabelCollection>
 171     </at:md>
 172   </entry>
 173 </feed>
 174 ```
 175
 176 Note how each `<t:TagLabel>` above has an `@xl:literalForm` which
 177 supplies the author‐provided tag label, while also labelling an R·D·F
 178 resource which gives the canonical tag. This is obviously quite verbose
 179 (by nature of being X·M·L), and more sophisticated solutions will
 180 doubtless be developed in the future, but it is already good enough to
 181 start building things on top of.
 182
 183 ## Pulling in Tags
 184
 185 Simply having the ability to fetch tags on request is likely
 186 insufficient for complex tagging systems or large repositories. A tag
 187 authority could have thousands of tags, and a repository needs to be
 188 able to not only discover them all but also keep them up·to·date
 189 without individually fetching each one. Inspired by the [<cite>I·I·I·F
 190 Change Discovery A·P·I</cite>][IIIF-Discovery], the current plan is to
 191 provide tag authorities with an [ActivityStreams][] event log which
 192 chronicles their changes, which repositories can then pull to update
 193 their models. The creation of a new romantic relationship tag
 194 `tag:ladys.example,2023-05-23:$BF-7607` might look like this :—
 195
 196 ```json
 197 { "@context": "https://ns.1024.gdn/Tagging/discovery.context.jsonld"
 198 , "type":
 199   [ "TagActivity"
 200   , "Create" ]
 201 , "context": "https://ladys.example/tag:ladys.example,2023-05-23:"
 202 , "object": "https://ladys.example/tag:ladys.example,2023-05-23:$BF-7607"
 203 , "endTime": "2023-05-23T19:33:33-07:00"
 204 , "label": "Add Shadow ❤ O·C tag"
 205 , "comment": "This is the romantic equivalent of the existing (sexual) Shadow × O·C tag."
 206 , "states":
 207   [ { "predicate": "a"
 208     , "object": "RomanticRelationshipTag" }
 209   , { "predicate": "prefLabel"
 210     , "object":
 211       { "@value": "Shadow ❤ O·C"
 212       , "@language": "en" } }
 213   , { "predicate": "involves"
 214     , "object": "https://ladys.example/tag:ladys.example,2023-05-23:N54-M773" }
 215   , { "predicate": "involves"
 216     , "object": "https://ladys.example/tag:ladys.example,2023-05-23:V2G-36D8" } ] }
 217 ```
 218
 219 A change in the preferred label of that tag might look as follows :—
 220
 221 ```json
 222 { "@context": "https://ns.1024.gdn/Tagging/discovery.context.jsonld"
 223 , "type":
 224   [ "TagActivity"
 225   , "Update" ]
 226 , "context": "https://ladys.example/tag:ladys.example,2023-05-23:"
 227 , "object": "https://ladys.example/tag:ladys.example,2023-05-23:$BF-7607"
 228 , "endTime": "2023-05-23T19:34:20-07:00"
 229 , "label": "Add variation selector 15 after heart"
 230 , "comment": "Helps to prevent emojification on mobile platforms."
 231 , "unstates":
 232   [ { "predicate": "prefLabel"
 233     , "object":
 234       { "@value": "Shadow ❤ O·C"
 235       , "@language": "en" } } ]
 236 , "states":
 237   [ { "predicate": "prefLabel"
 238     , "object":
 239       { "@value": "Shadow ❤\uFE0E O·C"
 240       , "@language": "en" } } ] }
 241 ```
 242
 243 In each case note that the activity `unstates` and `states` various
 244 triples, with the `subject` of each assumed to be the same as the
 245 `object` of the activity. As these are published in an chronological,
 246 paginated list, it is easy for consumers to skip activities that they
 247 have seen before. And because each activity identifies as its `object`
 248 the tag being changed, they can easily skip activities for tags they
 249 don’t care about, too.
 250
 251 ## Implementation Plans
 252
 253 I’m currently working on a library for processing tag streams and a
 254 command‐line tool for generating them. Because this is a pull‐based
 255 model, it should be pretty easy to get a proof‐of‐concept up that is
 256 backed by static files and a minimal Caddy configuration. Then I will
 257 attempt to create a simple web viewer for browsing the tags, as well as
 258 start publishing works which use them. I’ll be sure to post updates to
 259 this blog when·ever any of this happens!
 260
 261 A lot of the initial discussion for this work has taken place on the
 262 [Fandom Coders][Fancoders] Discord server. Shoutout to the people there
 263 for putting up with my many 4 A·M Freeform diagrams and rants about
 264 ActivityStreams.
 265
 266 [AO3]: <https://archiveofourown.org/> "Archive of Our Own"
 267 [ActivityStreams]: <https://www.w3.org/TR/activitystreams-core/> "ActivityStreams 2.0"
 268 [Fancoders]: <https://www.fancoders.com> "Fandom Coders"
 269 [IIIF-Discovery]: <https://iiif.io/api/discovery/1.0/> "IIIF Change Discovery API 1.0"
 270 [JSON-LD]: <https://json-ld.org> "JSON‐LD: JSON for Linking Data"
 271 [OTW]: <https://www.transformativeworks.org/> "Organization for Transformative Works"
 272 [RFC4287]: <https://www.rfc-editor.org/rfc/rfc4287.html> "The Atom Syndication Format"
 273 [Tagging]: <https://ns.1024.gdn/Tagging/> "The Tagging Vocabulary"
 274 [draft-nottingham-atomtriples-00]: <https://datatracker.ietf.org/doc/html/draft-nottingham-atomtriples-00> "AtomTriples: Embedding RDF Statements in Atom"
 275 [end_otw_racism]: <https://blog.ladys.computer/2023-05-14/end_otw_racism/> "End Racism in the O·T·W"
 276 [satsuma:25796]: <https://satsuma.dreamwidth.org/25796.html> "A Chronic Habit of Avoiding Responsibility? #EndOTWRacism"
 277 ]]></sioc:content>
 278         <dc11:rights rdf:parseType="Markdown"><![CDATA[
 279 Copyright © 2023
 280 <a href="https://www.ladys.computer/about/#lady">Lady</a>
 281 <small>[ActivityStreams Hatefucker]</small>.
 282 Some rights reserved.
 283
 284 This blogpost is licensed under a <a rel="license"
 285 href="http://creativecommons.org/licenses/by/4.0/"><cite>Creative
 286 Commons Attribution 4.0 International License</cite></a>.
 287 ]]></dc11:rights>
 288 </awol:Entry>