Re: Follow/Following/follows

On 04/22/2015 06:33 PM, Breslin, John wrote:
> Hi all
Hi John o/

> 
> Sorry I was muted and couldn't answer you in the call Ann. (My fault for creating feedback between landline and mobile as I transitioned between phones.) I tried downloading IRC on mobile but then my signal went and that was the end of that. 
We appreciate that you took all the effort of trying to still
participate in this call!

I also hope we can continue next days in async with github issues,
mailing list and wiki. Always IRC also provides nice sync communication
channel which we can upgrade to audio or even video whenever needed.

> 
> I found elf's proposal interesting and will send some more comments later. It nicely addresses an issue with static dumps (we've had in SIOC) that show follows or moderates or subscribes relationships which are specific to that point in time (of the dump) but don't reflect the full timeline.
Happy to hear that and thank you John for binging your real word
experience from SIOC work!

I would like to put emphasis once again that I just propose rough
framework to discuss all those topics further, not so much propose any
concrete vocabulary approach. Also patterns I try to understand and
reuse come from existing and established sources.
* http://www.w3.org/TR/rdf-schema/#ch_reificationvocab
* http://www.w3.org/TR/swbp-n-aryRelations/
* http://patterns.dataincubator.org/book/

I still need to figure out how CCO uses cco:habit and cco:topic together
with owl:propertyChainAxiom. Also before F2F I plan to publish some more
examples of cco:skill, cco:expertise and cco:interest via my WebID (well
almost) https://wwelves.org/perpetual-tripper (broken RDFa but also CN
application/ld+json and turtle will come soon!) It will also come useful
for the Open Badges / Web Credentials and xAPI work involving claims and
endorsements.



> I think the names of terms are still an issue as you really have to think hard about what each means, and following as a verb and "a following" as a noun may not be intuitive at first. But once there is a pattern it may be more obvious. 
I agree with that and currently just experiment with naming...

schema.org or uses FooAction pattern
* http://schema.org/docs/full.html#Action

yesterday I thought about using different prefixes e.g.
* rel:follows
* qualified:Follows
* action:Follow
* action:Join
* activity:Listen
* operation:Create (HTTP POST)
* action:Create (subClassOf action:Cook, action:Grow)
* rel:attendee
* rel:member
* qualified:Membership
* qualified:Attendance
* action:Offer
* rel:mentor
* qualified:Mentorship


> 
> Am I right in describing it as below?
> 
> What causes it - action - follow
> What is the process - activity - following that starts and ends?
> What is the result - a relationship created - follows or followed (we use the former because of convention on social sites but the latter may may sense).
Yes! Graph diagrams which I put in Activity subtypes visualize it e.g.
* https://github.com/w3c-social/social-vocab/tree/master/activity/Follow
* https://github.com/w3c-social/social-vocab/tree/master/activity/Join
* https://github.com/w3c-social/social-vocab/tree/master/activity/Travel

Besides creating unqualified *follows* (similar to foaf:knows) it can
also add qualified relation which I tried calling *Following* and
associate Activity/Action which caused it  (eg. :interaction). After
*Unfollow* action, it will remain also having it associated, while
*follows* edge will disappear.

I need to turn it into slides for different states like before and after
Action/Activity/Operation. e.g.

# graph *before* action:Follow
<https:/rhiaro.co.uk> a :Person ;
  :name "Amy Guy" .

<https://wwelves.org/perpetual-tripper> a :Person ;
  :name "elf Pavlik* ;


# graph *after* action:Follow
<https:/rhiaro.co.uk> a :Person ;
  :name "Amy Guy" .

<https://wwelves.org/perpetual-tripper> a :Person ;
  :name "elf Pavlik* ;
  :follows <https://rhiaro.co.uk> . # *<-*

<https://wwelves.org/perpetual-tripper/rel/3242> a :Following ;
  rdf:subject <https://wwelves.org/perpetual-tripper> ;
  rdf:predicate :follows ;
  rdf:object <https:/rhiaro.co.uk> ;
  :interaction <https://wwelves.org/perpetual-tripper/log/2762> ;
  :startTime "16 July 2014" .

<https://wwelves.org/perpetual-tripper/log/2762> a :Follow ;
  as:actor `<https://wwelves.org/perpetual-tripper> ;
  as:object <https:/rhiaro.co.uk> ;
  as:result `<https://wwelves.org/perpetual-tripper/rel/3242> .

#graph *after* action:Unfollow
<https:/rhiaro.co.uk> a :Person ;
  :name "Amy Guy" .

<https://wwelves.org/perpetual-tripper> a :Person ;
  :name "elf Pavlik* ;
  # *no more* :follows

<https://wwelves.org/perpetual-tripper/rel/3242> a :Following ;
  rdf:subject <https://wwelves.org/perpetual-tripper> ;
  rdf:predicate :follows ;
  rdf:object <https:/rhiaro.co.uk> ;
  :interaction <https://wwelves.org/perpetual-tripper/log/2762> ,
    <https://wwelves.org/perpetual-tripper/log/12310> ;
  :startTime "16 July 2014" ;
  :endTime "30 April 2015" . # *<-*

<https://wwelves.org/perpetual-tripper/log/2762> a :Follow ;
  as:actor `<https://wwelves.org/perpetual-tripper> ;
  as:object <https:/rhiaro.co.uk> ;
  as:result `<https://wwelves.org/perpetual-tripper/rel/3242> . #
creation of first 'revision'

<https://wwelves.org/perpetual-tripper/log/12310> a :Unfollow ;
  as:actor `<https://wwelves.org/perpetual-tripper> ;
  as:object <https:/rhiaro.co.uk> ;
  as:result `<https://wwelves.org/perpetual-tripper/rel/3242> . # more
operation:Update on that resource, or its newer revision?

besides slides i could also imagine nice diff like in
* https://gist.github.com/elf-pavlik/029917ccc535e889f693/revisions

BTW let's move further conversation off mailing list to github so we can
use nicer code snippets? https://github.com/w3c-social/social-vocab/issues

> 
> Same for:
> 
> Subscribe
> Subscribing
> Subscribes/Subscribed
> 
> I realise my summary with actions and activities may confuse with ActivityStreams terms but it helps me understand. 
I don't know if you had chance to look at current AS2.0 working drafts
but what you say makes sense to me!

I would also bring to our attention Action, Operation and Activity
distinction.
1) How they differ in sense conceptual interactions between agents and
objects vs. API calls on web resources  (maybe even connected to
httpRange-14 in a way?)
2) How they differ in affordance / potential action vs. ongoing or past
activity.

To my understanding Linked Data Platform and Hydra API focus on
Operation - interacting with web resources. Schema.org Potential Action
and as far as I remember AS2.0 Action Handlers kind of mix it with
Action in conceptual sense - Join, CheckIn, Play as well as Activity
Listen, Travel etc.

I already shared couple of times my concerns on public-vocabs that
Schema.org Potential Action currently may not define it clearly e.g.
http://schema.org/error (ReadAction failed because I didn't bring my
glasses or because of HTTP 404?)


Once again, thank you for joining efforts & cheers!


> 
> John 
>

Received on Wednesday, 22 April 2015 17:48:56 UTC