Helsinki Data Week reposted this
Many have asked, here they are: My slides form Helsinki Data Week fun fact: We will keep the current title. Searching for "Meta" on O'Reilly's learning platform returns search results about a specific tech company. Oh well If you use diagrams from these slides, please refer me as the source. Antti Rask 🦜Juha Korpela Eevamaija Virtanen Säde Haveri #metagrid #metadatamanagement #metadata
I think I agree with Philippe NIEUWBOURG, in the sense that I think you have exposed the problem very well. Dare I say it but the (current) solution is staring us in the face. It is spreadsheets, some up to date, some out of date - it is also wiki sites, like Confluence, Notion, also in various stages of currecny - and it is people, for people are the great interpreters for our IT landscape. Whether I work in a global e-commerce giant that employs 1.5 million people, or a large local telco with 50,000 employees, or a small innovative startup with 20 to 30 people the problem of searching for data always involves people at some point. I don't see people in your metagrid. I'm having a quick skim of your pre-release book and I begin to see your problem and your thesis. One of the things I love about Zhamak's version of Data Mesh is that she clearly established the socio-technical nature of it. She recently clarified that it was social AND technical. I'd be tempted to add ecological to the mix but I am afraid that I may make it into a word salad.
Martin Holevang Sandra Anastasia N. Kristholm har I set den her?
I have some thoughts on the story part of this story. Happy to chat. Did you present this live somewhere? Also I vote for “meta DATA grid” as a title otherwise people might think it’s about a grid about a grid!😉
Seems very interesting but you lost me at slide 8. I think you exposed very well the problem. And we understand it. The solution seems to me more confusing. Sure I need to read the book 😂
Where and how ontologies and shacl contraints fit in ? Slide 5 without explicit definition can mean many things, for me personally BMW, Citroen, etc. represent: - instances of class schema:Corporation - instances of car corporations - ... where a car, an example/instances, e.g. Tesla Y can be defined as knowledge graph as such: :TeslaY a :schema:Car :TeslaY pav:createdBy :Tesla :Tesla a :CarCorporation, prov:Agent :CarCorporation rdfs:subClassOf schema:Corporation Therefore, Is metagrid knowledge graph ? i.e. semantic data model (shacl + owl) + instances? or something completely new?
Thank you for sharing! Here are my thoughts after looking at the slide. In the Edit mode/pane, there's a default index established by the order of the slides. However, in the View mode that doesn't translate into slide numbers. I'm trying to make a parallel to the meta grid, considering that we have an implicit order in the left, but no way of knowing the order on the right, unless we explicitly include it (slide #s). The information is there. Should we also be able to quickly know where/how it fits? I think we do. Thoughts?
Thanks for sharing. Diving into this topic towards the end of the year 👍
🐙
Executive Product Owner | Transforming Data into Impact: Driving Innovation in Mobility with Passion and Precision
1moHey dbt Labs, I've just seen this exciting article on the subject of Meta Grid. The way metadata is handled in dbt should make it easy to use across domains, right? You already have the foundation for building a meta grid in your application.