IRC log of rdf-star on 2024-04-11

Timestamps are in UTC.

15:32:44 [RRSAgent]
RRSAgent has joined #rdf-star
15:32:48 [RRSAgent]
logging to https://www.w3.org/2024/04/11-rdf-star-irc
15:32:48 [Zakim]
RRSAgent, make logs Public
15:32:49 [Zakim]
please title this meeting ("meeting: ..."), pchampin
15:33:12 [pchampin]
meeting: RDF-star WG biweekly focused meeting
15:33:42 [pchampin]
pchampin has changed the topic to: RDF-Star WG - Focused meeting - 2024-04-11
15:34:02 [pchampin]
Regrets+ olaf, gwilliams, azimmermann, draggett, fsasaki
15:34:10 [pchampin]
present+ ktk
15:34:13 [pchampin]
present+ pchampin
15:55:52 [TallTed]
TallTed has joined #rdf-star
15:57:16 [fsasaki]
fsasaki has joined #rdf-star
15:57:34 [pfps]
pfps has joined #rdf-star
15:58:23 [gkellogg]
gkellogg has joined #rdf-star
15:58:42 [niklasl]
niklasl has joined #rdf-star
15:59:06 [TallTed]
TallTed has changed the topic to: RDF-Star WG — Focused meeting — 2024-04-11 agenda: https://www.w3.org/events/meetings/5ecc5c5f-5cd2-410c-b97c-6b13c6b843f1/20240411T120000/
15:59:12 [TallTed]
RRSAgent, draft minutes
15:59:13 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html TallTed
15:59:32 [TallTed]
present+
15:59:53 [fsasaki]
present+
15:59:57 [gkellogg]
present+
16:00:32 [TallTed]
previous meeting: https://www.w3.org/2024/04/05-rdf-star-minutes.html
16:00:34 [TallTed]
next meeting: https://www.w3.org/2024/04/12-rdf-star-minutes.html
16:00:36 [TallTed]
agenda: https://www.w3.org/events/meetings/5ecc5c5f-5cd2-410c-b97c-6b13c6b843f1/20240411T120000/
16:00:36 [agendabot]
clear agenda
16:00:36 [agendabot]
agenda+ Discuss if a single id can reify more than one triple
16:00:43 [ora]
ora has joined #rdf-star
16:00:51 [tl]
tl has joined #rdf-star
16:01:11 [AndyS]
present+
16:01:14 [TallTed]
RRSAgent, draft minutes
16:01:15 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html TallTed
16:01:27 [tl]
present+
16:02:00 [TallTed]
scribe: ?scribe?
16:02:02 [TallTed]
chair: ora
16:02:14 [ora]
present+
16:03:25 [doerthe]
doerthe has joined #rdf-star
16:03:29 [AndyS]
scribe: AndyS
16:03:30 [TallTed]
s/?scribe?/ AndyS
16:03:31 [pfps]
present+
16:03:37 [doerthe]
present+
16:03:47 [TallTed]
RRSAgent, draft minutes
16:03:49 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html TallTed
16:04:03 [niklasl]
present+
16:04:05 [TallTed]
Zakim, next item
16:04:05 [Zakim]
agendum 1 -- Discuss if a single id can reify more than one triple -- taken up [from agendabot]
16:04:25 [AndyS]
ora: The focus of today is continue the discussions on reification : many-to-one, many-to-many.
16:05:14 [AndyS]
... AWS Neptune team discussions, not telling RDF-star WG what to do
16:05:46 [AndyS]
... Neptune focus on interoperability between LPG and RDF data.
16:05:49 [pfps]
q+
16:06:39 [AndyS]
... can't understand the extended scenario helps given customer discussions.
16:06:41 [Souri]
Souri has joined #rdf-star
16:06:43 [ora]
ack pfps
16:06:47 [Souri]
present+
16:06:58 [enrico]
enrico has joined #rdf-star
16:07:02 [enrico]
present+
16:07:03 [enrico]
q+
16:07:05 [pfps]
sorry, mike problems
16:07:14 [ora]
ack enrico
16:07:24 [pfps]
q-
16:07:32 [pfps]
q+, maybe mike problems are fixed
16:07:37 [AndyS]
enrico: what is the danger of the extended case
16:07:59 [AndyS]
... can it be that advice is to use many-to-one?
16:08:39 [AndyS]
... what is the damage of allowing many-to-many?
16:08:55 [tl]
q+
16:09:15 [AndyS]
ora: your (Enricos) case is allow it and advise on its usage.
16:10:12 [AndyS]
ora: hard for vendors of triplestores -- not a complete implementation
16:10:29 [Souri]
q+
16:10:50 [AndyS]
... for LPG, hard to explain that a property covers "multiple edges"
16:11:24 [AndyS]
enrico: profiles would be an approach but we end up with too many profiles
16:11:30 [TallTed]
q+
16:11:52 [AndyS]
ora: well-formedness is like lists or classic reification case today
16:12:08 [ora]
ack tl
16:12:09 [AndyS]
q+
16:12:10 [enrico]
q+
16:12:42 [AndyS]
tl: two things: Q1: how is many-to-many not possible in LPG?
16:13:12 [fsasaki]
q+
16:13:28 [AndyS]
ora: we didn't find it intuitive in our group.
16:13:52 [pchampin]
intuitive is by definition different for different people
16:13:54 [AndyS]
tl: annotation syntax is optimized many-to-one case
16:14:15 [AndyS]
s/many/for the many/
16:14:19 [niklasl]
+1 to pchampin re. intutition
16:14:44 [TallTed]
q?
16:14:59 [ora]
ack Souri
16:15:16 [AndyS]
ora: not an engineering question - it is the intuitive understanding
16:15:59 [AndyS]
souri: interoperability of LPG/RDF1.2 -- since LPG have edge properties with only values
16:16:33 [AndyS]
... in RDF 1.2 is general - - edge on edge
16:16:47 [pchampin]
+1 to what Souri says
16:17:00 [AndyS]
... needs to be considered when converting to LPG
16:17:35 [ora]
q+
16:18:02 [gkellogg]
present+ Kurt_Cagle
16:18:09 [AndyS]
... w do not want to RDF 1.2 because of an LPG restriction so consider N-to-1 and N-to-N
16:18:23 [AndyS]
s/w do/we do/
16:18:26 [ora]
ack TallTed
16:18:48 [pchampin]
and that does not prevent us from providing guidelines for "reading" (different subsets of) RDG 1.2 as LPG
16:19:12 [AndyS]
TallTed: The depth of your (Ora) feeling suggests you want the restriction.
16:19:31 [pchampin]
q+
16:19:44 [AndyS]
... N-to-1 breaks RDF principles in a new way.
16:19:50 [pchampin]
q-
16:19:55 [AndyS]
ora: this is about well-formedness
16:20:06 [gkellogg]
q+
16:20:12 [AndyS]
tallted: that's your right
16:20:36 [AndyS]
... one reifier - multi terms is the next logical step
16:20:36 [Dominik_T]
Dominik_T has joined #rdf-star
16:20:39 [ora]
ack AndyS
16:20:41 [pchampin]
q+
16:20:51 [gkellogg]
scribe+
16:20:52 [pfps]
Well-formedness was introduced, I think, to get around a problem with merging graphs, nothing more
16:20:54 [pfps]
q+
16:20:55 [Dominik_T]
present+
16:21:17 [gkellogg]
AndyS: It's not quite so simple to discuss the differences between systems, we have something called the "Semantic Way".
16:21:40 [ora]
ack enrico
16:21:45 [gkellogg]
... If you make restrictions you can impede how different systems interact.
16:21:49 [gkellogg]
scribe-
16:22:23 [enrico]
<< :w1 | :liz :married :richard >> :location :las-vegas .
16:22:27 [AndyS]
enrico: technical: in the LPG standard there are undirected edges
16:22:30 [enrico]
<< :w1 | :richard :married :liz >> :date 1966 .
16:22:57 [AndyS]
... this means we need N-to-M
16:24:17 [AndyS]
... my other argument is why restrict the well-formedness fragment to "to-one" . We need many well-formed.
16:25:16 [AndyS]
... 3rd argument: open world means two URIs can denote the same resource hence n-ary case [scribe: i.e merge]
16:25:27 [pchampin]
q?
16:25:33 [Kurt]
Kurt has joined #rdf-star
16:25:39 [enrico]
:e rdf:reifies <<( :s1 :p1 :o1 )>> .
16:25:44 [enrico]
:e rdf:reifies <<( :s2 :p2 :o2 )>> .
16:25:53 [enrico]
ASK WHERE { _:x rdf:reifies <<( :s1 _:y :o2 )>> } ==> TRUE
16:26:10 [ora]
ack enrico
16:26:35 [enrico]
[] a ex:Edge ;
16:26:35 [enrico]
:on "2023"^^xsd:gYear ;
16:26:35 [enrico]
rdf:reifies <<(<book> :publisher <house>)>> .
16:26:55 [enrico]
ex:Edge a owl:Class ;
16:26:55 [enrico]
rdfs:subClass [ a owl:Restriction;
16:26:55 [enrico]
owl:onProperty rdf:reifies ;
16:26:55 [enrico]
owl:cardinality 1 ] .
16:27:11 [gkellogg]
s/[] a ex:Edge ;//
16:27:19 [gkellogg]
s/ :on "2023"^^xsd:gYear ;//
16:27:19 [AndyS]
enrico: remove first 3 triples at "a ex:Edge"
16:27:30 [gkellogg]
s/ rdf:reifies <<(<book> :publisher <house>)>> .//
16:28:08 [AndyS]
enrico: rdf:reifies is not functional in RDF.
16:28:23 [AndyS]
... this should be covered in the best practices
16:28:28 [ora]
ack fsasaki
16:28:45 [ora]
q-
16:29:12 [AndyS]
fsasaki: I undertand the AWS Neptune feedback was user based. Similar in SAP.
16:29:43 [AndyS]
... in LPG framing, easier to explain currently.
16:29:50 [ora]
q+
16:29:57 [AndyS]
... so this is an opportunity for RDF
16:30:28 [AndyS]
... the price is less features - clarity of one case over fully generality
16:30:37 [AndyS]
... user groups are AI users
16:30:42 [ora]
ack gkellogg
16:30:58 [doerthe]
only way out here would be referential opacity (just saying, I really do not want to get back to that discussion ;) )
16:31:03 [AndyS]
gkellogg: well-formedness well discussed
16:31:36 [AndyS]
... none defined by RDF vocabulary - "semantic extensions may impose limits"
16:31:36 [pchampin]
i like the "less features can mean more adoption" approach
16:32:06 [AndyS]
... we could have examples with only allowing literals via annotations
16:32:29 [AndyS]
... many-to-one does not remove the UCs - suggests using a container
16:32:34 [ora]
ack pchampin
16:33:33 [AndyS]
pchampin: With tallted about say anything about anything - not seeing our choice as "illegal"
16:33:36 [fsasaki]
q+
16:34:15 [AndyS]
... rdf terms same-term, same value example -- 042 and 42
16:34:28 [AndyS]
.. across various stores
16:34:49 [AndyS]
s/.. across/... across/
16:35:19 [ora]
ack pfps
16:35:31 [AndyS]
... keep RDF open and better at expressing implementation choices
16:35:48 [AndyS]
pfps: I agree with the "many" people.
16:35:54 [tl]
+1 to pchampin
16:36:15 [ora]
ack ora
16:36:19 [AndyS]
... restricting RDF does not work for me. May be other reasons. I haven't heard them.
16:37:05 [AndyS]
ora: Worked with RDF for a long term - found when LPG emerged it was more work
16:37:48 [AndyS]
... so why do people choose LPG? Several reasons - one is the model is easier to understand
16:37:57 [tl]
q+
16:38:12 [AndyS]
... struggle to keep RDF relevant in the industry
16:38:18 [pfps]
I am again puzzled as to why RDF should be limited just because LPGs are becoming more popular.
16:38:39 [TallTed]
RRSAgent, draft minutes
16:38:41 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html TallTed
16:39:14 [AndyS]
... we see align as a way to keep RDF relevant
16:39:32 [enrico]
q+
16:39:57 [TallTed]
q?
16:40:05 [AndyS]
... focus on a solution causes more adoption
16:40:07 [ora]
ack fsasaki
16:40:11 [TallTed]
q+
16:40:26 [pfps]
The argument that RDF should be limited to what is in LPQ sounds like "we have to destroy RDF in order to save it"
16:40:33 [fsasaki]
https://www.deeplearning.ai/short-courses/knowledge-graphs-rag/
16:40:42 [AndyS]
fsasaki: not AWS vs group. SAP has the same experience in the AI area
16:41:41 [AndyS]
... disagree about profiles. e.g. xHTML. Limited uptake.
16:41:44 [pchampin]
q+
16:41:52 [ora]
ack tl
16:42:25 [AndyS]
tl: OWL profiles work quite well. Main complexity in RDF comes from blank nodes.
16:42:46 [AndyS]
... because open world which is a key feature.
16:43:04 [ora]
q+
16:43:36 [AndyS]
... easier to work/model in a known domain
16:44:09 [TallTed]
q+ adrian
16:44:45 [AndyS]
... profiles done right would be a good idea
16:45:45 [pfps]
The trouble with the strict division into edges and attributes in LPG is that you can't then, for example, say where Dick and Liz were married by putting an attribute on a marriage edge. Well, you can, and I've seen it done, by using "strings for things", which is a very bad idea.
16:46:08 [fsasaki]
q+
16:46:22 [ora]
ack enrico
16:46:30 [fsasaki]
q-
16:46:52 [fsasaki]
q+
16:47:26 [Kurt]
q+
16:47:32 [AndyS]
enrico: to be constructive: (1) well-formedness for many-to-one (2) several profiles (not a good choice IMO) - considerable work (3) have a best practices section.
16:47:38 [Kurt]
q-
16:48:00 [pfps]
I'm also confused as to just what the status of well-formedness is going to be. Even if the WG says that well-formedness includes many-to-one reifiers, systems should support the many-to-many case.
16:48:09 [AndyS]
... keep many-to--many and in the well formed explanation has an interpretation.
16:49:17 [AndyS]
s/to--many/to-many/
16:49:33 [ora]
ack TallTed
16:49:46 [AndyS]
.. BP choice -- reificiations always have a meaning
16:49:56 [AndyS]
s/.. BP/... BP/
16:50:01 [Kurt]
q+
16:50:17 [niklasl]
+1 (explain, using good, intuitive real-world(ish) examples, different kinds of reification)
16:50:35 [AndyS]
tallted: LPGs appeal as fewer rules.
16:51:43 [AndyS]
... make choices and may have consequences later
16:51:58 [tl]
I would be totally okay with best practices favoring many-to-one edge annotations, and, as I said, the annotation syntax favors them already
16:52:23 [Kurt]
LPGs (specifically Neo4J) is also is an abstraction layer on top of a core model - RDF is closer to "machine language" to most people.
16:52:25 [Souri]
Can we create an "LPG profile of RDF1.2"? This profile would restrict RDF1.2 in every way needed for keeping it limited to what LPG can support.
16:53:47 [niklasl]
Yes (to tl). And to Souri, we can at least have an LPG best practises section (also feeding from what Ted on "if you hit a limit, migrate" + [my addition] "to RDF" ;) ).
16:54:44 [pfps]
SPARQL (also SQL) is complex indeed. But I find the hardest to understand complexity is where the bottom-up execution model is violated.
16:55:07 [tl]
+1 to an "LPG profile of RDF1.2"
16:55:43 [ora]
ack pchampin
16:56:37 [AndyS]
pchampin: about profiles. They such things already (not called "profile").
16:57:09 [AndyS]
... do not believe well-formedness is being used the same way by each of us.
16:57:15 [TallTed]
in my experience, SPARQL's "bottom up" is too often heard/read as "from the bottom line of the written query to the top ilne" instead of "from the deepest inner sub query, to the highest outer super query"
16:57:26 [TallTed]
hence, my calling it "inside-out"
16:57:35 [ora]
ack ora
16:58:21 [AndyS]
ora: Like the idea of best practices.
16:58:35 [enrico]
Answer to Pierre-Antoine: this is the formal definition of well formedness discussed and agreed upon so far: https://github.com/w3c/rdf-star-wg/wiki/RDF%E2%80%90star-semantics%3A-option-3#best-practices-for-reification-and-reification-well-formedness
16:58:48 [ora]
ack adrian
16:58:52 [AndyS]
... how is that going to work if imples exlude non-best-practice
17:00:06 [AndyS]
adrian: in last 10 years, we restrict ourselves for an application by emphasises open-world / the long term.
17:00:35 [ora]
ack fsasaki
17:00:38 [AndyS]
... hope we don't break open world .. need to restrict complexity because it confuses people.
17:01:05 [ora]
ack Kurt
17:01:08 [AndyS]
fsasaki: any consumer can take any producer data.
17:01:24 [pfps]
There is a bit in 1984 that I think is relevant here: Two ministries are feuding, with the metric being how much floor space they have relative to the other. In the meantime, a third ministry is increasing its floor space at both their expenses.
17:02:10 [fsasaki]
s/producer data/producer data, that feature of RDF may be harmed if one data producer uses profile 1 and the other profile 2 etc./
17:02:18 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html fsasaki
17:02:36 [AndyS]
Kurt: LPGs encodes complexity into bundles
17:02:44 [AndyS]
... then work with bundles
17:02:57 [AndyS]
... RDF is a substrate language.
17:03:44 [AndyS]
... abstraction layers are what users are attracted to.
17:04:33 [pchampin]
rrsagent, make minutes
17:04:34 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html pchampin
17:05:07 [gkellogg]
regrets- fsasaki
17:05:09 [AndyS]
SemTF tomorrow
17:05:13 [gkellogg]
present+ fsasaki
17:06:18 [gkellogg]
rrsagent, make minutes
17:06:19 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html gkellogg
17:45:49 [gkellogg]
gkellogg has joined #rdf-star
18:10:00 [gkellogg]
gkellogg has joined #rdf-star
18:26:19 [gkellogg]
gkellogg has joined #rdf-star
18:46:56 [gkellogg]
gkellogg has joined #rdf-star
19:00:56 [TallTed]
Zakim, end meeting
19:00:56 [Zakim]
As of this point the attendees have been ktk, pchampin, TallTed, fsasaki, gkellogg, AndyS, tl, ora, pfps, doerthe, niklasl, Souri, enrico, Kurt_Cagle, Dominik_T
19:00:59 [Zakim]
RRSAgent, please draft minutes
19:01:00 [RRSAgent]
I have made the request to generate https://www.w3.org/2024/04/11-rdf-star-minutes.html Zakim
19:01:05 [Zakim]
I am happy to have been of service, TallTed; please remember to excuse RRSAgent. Goodbye
19:01:05 [Zakim]
Zakim has left #rdf-star
19:01:08 [TallTed]
RRSAgent, bye
19:01:08 [RRSAgent]
I see no action items