Lmst

Querying your SQL database in natural language?

Simone Pandini and Manuela Corte Pause will show how this can be implemented by using open source software based on semantic technologies.

@ontop
#rdf
#KnowledgeGraphs
#llm

A bit of good news. Saving and restoring from RDFXML now preserves the blank node identifiers.

I can now simplify the next step of the microplanner test data.

The basic plan was for the process prior to the microplanner should create a model containing lists of RDF statements that would be formed into a paragraph. It was beginning to look like that model would have to also contain those statements copied from other models. Now it can just reference those models (as submodels). Reification will work across models.

There is still the problem of using submodels with large RDF models. It works fine for small ones, but things like WordNet are a bit of a disaster.

3/2

#SASSY #RDF

i asked me, how i can create an #cid of an #rdf file, that represents properties of one subject. because idealy the subject contains the #cid. or are i'm wrong ?
#ipfs

What makes the research fun is that the Redland libraries use the C version of polymorphism: Structs full of function pointers. You have to hope that you can find where, and when, they were initialised and that they don't get changed somewhere.

The good news is that after following numerous false trails I found a spot that can be messed with to defeat the overwriting.

I will see how this turns out....

2/2

#SASSY #RDF #programming

I have just been for a dive into the Redland RDF libraries. I have been developing a C++ wrapper for them and noticed something which seems odd, and which would defeat something I am planning to do with it.

RDF has the option of using anonymous or blank nodes (along with URI based nodes, and literal values). These are convenient as they remove the need to name everything (the hardest problem in programming).

If you create a blank node with an identifier it uses that, otherwise it generates an identifier itself. If you are saving to a PostgreSQL database that would be the end of the story. However if you save to an XML file, which is convenient, it overwrites your identifier with a generated one during the reading of the XML. These ids never repeat, even for the same node, between openings of the file. The behaviour is quite different to the database storage.

My intention is to use UUIDs for the blank node identifiers to give them a permanent identity. This will allow models to reference all nodes in other models which will make combining models into larger ones possible.

1/2

#SASSY #RDF

The more things change, the more they stay the same: #NLWEB ( https://news.microsoft.com/source/features/company-news/introducing-nlweb-bringing-conversational-interfaces-directly-to-the-web/ ) strikes me as a direct descendent of the aspirations built into #RDF

#AIagents #AI

I've started a draft spec defining high level processing and rendering rules to bind an RDF graph to HTML elements.

The UI data binding is specified using HTML attributes, which when hydrated provides valid RDFa.

The idea is to enable bringing pages of data together within a single document

Still early days, but here's the description of a subject-connected list

https://jg10.solidcommunity.net/open-with/rdf-ui-binding.html#subject-connected-list

#LinkedData #RDF #RDFa #SolidProject #PodOS

Many new Ontopic Studio users come from Stanford's Protégé open-source ontology editor.

While using Protégé, together with the Ontop plugin, gives excellent start to virtual knowledge graphs and knowledge graph construction, there are a couple of things users should be aware of, when start using Ontopic Studio. We discuss them explicitly in our documentation section.

https://docs.ontopic.ai/studio/learn/coming-from-protege.html

#KnowledgeGraphs #rdf

Practical exercise of the day with #TaganakNet, the #RDF #Rust SDK I am working on with @codecraft :

I have noted, in a Turtle file, that I own some albums and songs in my multimedia library. I have noted that as #IPFS IRIs linked to the #MusicBrainz entities of the songs and albums.

A small CLI tool reads the data as RDF graph, links them with #LinkedData from MusicBrainz, and generates a working M3U playlist using my IPFS gateway.

Screenshot of a snippet of Turtle data:

@prefix wd: <http://www.wikidata.org/entity/> .
@prefix mbrec: <http://musicbrainz.org/recording/> .
@prefix mbrel: <http://musicbrainz.org/release/> .
@prefix sf: <http://slide.fish/vocab#> .
@prefix ma: <https://www.w3.org/ns/ma-ont#> .
@prefix dc: <http://purl.org/dc/terms/> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .

<> a sf:Collection ;
sf:accessProfile [
sf:ipfsGateway <https://koloni.squirrel.land/> ;
] ;
sf:peerCollection <./filme.ttl> ;
rdfs:seeAlso <https://codeberg.org/Natureshadow/slidefish-collection/raw/musik.ttl> .

mbrel:88979e59-6c00-4d77-ad4c-1d9bfe8a305d a sf:Album ;
sf:Collection <> ;
dc:title "Tierparklieder aus Olderdissen" ;
rdfs:seeAlso wd:Q134531885 .

mbrec:664bb18a-995a-4179-a1c4-9508e406e212 a sf:Song ;
sf:collection <> ;
sf:media <ipfs://bafybeigim4smrbcjz5ehyqwmy5aprjf4s4qxn7uqhfd4eg37nhibo4nkaq> .

mbrec:6a420b9a-4fc2-446a-bf26-dec109e357ca a sf:Song ;
sf:collection <> ;
sf:media <ipfs://bafybeicsaylu5bw3itfkg2tmuzmv4yvo3t7sjgm6x4b3eiqy6eokqu2eka> .

Screenshot of a command line:

cargo run -- file:///home/nik/frifot-kopie/Privat/hoernchen-hirn/slidefish-collection/musik.ttl music album http://musicbrainz.org/release/88979e59-6c00-4d77-ad4c-1d9bfe8a305d m3u

and VLC media player playing the album "Randale – Tierparklieder aus Olderdissen"

I have just released my first Zed extension!

This is a RDF Turtle format extension that features a LSP, highlighting, and symbol navigation.

Read about it here : http://moskitohero.com/zed-turtle-extension/

#zed #ttl #rdf #rust

Is there any activity regarding #rdf databases and #ipfs? Ideally openSource ;-)

A new tender for Enhancement of the EU Public Procurement Data Space, 4.8 mil https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/tender-details/1837c208-7869-4190-bf65-3bccd8a1f919-CN

#RDF #SPARQL #knowledgegraph #dataspace

[just found a nice RDF blog by Bob DuCharme:]

"Reification is a red herring
And you don't need property graphs to assign data to individual relationships."

https://www.bobdc.com/blog/reification-is-a-red-herring/

#RDF #SemanticWeb

Anyone has an idea on why RDF 1.2 draft adds more logic and complexity to (the mostly unused & generally hated) reification?

At least, where's the substantiation?

#RDF #SemanticWeb

2/2

And of course you could always define an ontology using #RDF / #OWL to describe Agents. And I don't think it would be a big stretch to create an Agent Registry as an extension of #UDDI and maybe a dash of WSDL/S.

I haven't dug deeply enough into #A2A yet to really say, but I can't help but have a lurking hunch that maybe the FIPA standards (with maybe some small tweaks) would still be better and more general even to this day.

Inmiddels hebben vijf #provincies een officieel #Mastodon-account. Via onderstaande link kun je ze zien op een #openstreetmap-kaart, mogelijk gemaakt door #wikidata:

https://w.wiki/E4Ye

#overheid #linkeddata #rdf #osm #map #opendata #Nederland