Collection with members from result of SPARQL CONSTRUCT

Hi,

I currently do some work implementing Activity Streams 2.0 which we work
on in Social WG and recently core spec got published as FPWD[1].

I need a way to link to 'all my activity on some other thing', eg.

PREFIX hydra: <http://www.w3.org/ns/hydra/core#",>
PREFIX as: <http://www.w3.org/ns/activitystreams#>

CONSTRUCT { </some-collection> hydra:member ?activity }
WHERE {
  ?activity a as:Activity;
            as:actor https://wwelves.org/perpetual-tripper;
            as:object http://tmblr.co/ZxcvMv1UViUDg .
}

Preferably using a collection with permanent URI, eg.

{
  "@context": ...,
  "@id": "/some-collection",
  "@type": "Collection",
  "member: [
    {
      "@id": "/activities/1"
      "@type": ["as:Activity", "schema:DislikeAction"]
    },
    {
      "@id": "/activities/5"
      "@type": ["as:Activity", "schema:CommentAction"]
    },
    {
      "@id": "/activities/23"
      "@type": ["as:Activity", "schema:CommentAction"]
    }
  ]
}

(I do realize that in SPARQL example I didn't select those additional
types, etc.)

I wonder if hydra:manages or Linked Data Fragments could provide a way
to relate this collection to the query, from which its members come from?

In practice I would like to have a link in human UI, as well as
hypermedia control for machines in each resource of as:Activity type to
collection of all my other activities on that resource.

At some point I want to allow subscriptions to various collections of my
activities, each based on some filter/query. eg. { "@type":
"schema:TravelAction" }. BTW I also plan to setup LDF server for all the
data that i publish online, which will include very detailed log of my
activities, similar to http://aaronparecki.com/metrics :)


Thanks for any suggestions!


[1] http://www.w3.org/TR/2014/WD-activitystreams-core-20141023/

Received on Monday, 3 November 2014 16:20:32 UTC