Still patching brittle data pipelines?
Semantic models bring intent and intelligence to your data ecosystem; making your systems adaptable, transparent, and future-proof; with meaning, not just structure.
#SemanticAI #KnowledgeGraphs #ModelDriven #AI #DataEngineering #EnterpriseAI
19.06.2025 12:12
π 0
π 0
π¬ 0
π 0
"Move fast and break things" always sounded like a conman's mantra. Speed has its place, but without accountability, it's not innovation. It's recklessness, incompetence, even idiocy.
Real results come from planning and thoughtfulness, not just momentum.
29.05.2025 20:31
π 0
π 0
π¬ 0
π 0
πΏ π³ What if LLMs generated graph traversal paths via embeddings? This could break NP-hard limits in knowledge graphs, powering smarter automation, compliance, and AI copilots in regulated industries. LLMs may link symbolic logic with scalable graph computing. #AI #LLM #Automation #computerscience
19.05.2025 15:10
π 0
π 0
π¬ 0
π 0
ld-cli
π§ Developing with Linked Data, JSON-LD, CBOR-LD, or RDF/JCS Canonicalization? Try ld-cli π now available on Snap! π snap install ld-cli π
github.com/filip26/ld-cli
#LinkedData #JSONLD #CBORLD #RDF #JCS #RDFC #Canonicalization #CLI #OpenSource #DeveloperTools
11.05.2025 17:01
π 4
π 0
π¬ 0
π 0
Neurosymbolic hybrids have the potential to dramatically compress deep models by integrating logical reasoning and world models - leading to smarter, explainable, and more efficient AI.
20.04.2025 13:39
π 0
π 0
π¬ 0
π 0
π **Why Semantic Computing Over COBOL?**
1. Semantic models (e.g., Datalog, F-Logic) handle complex logic declaratively
2. Easier to adapt to new requirements and business rules
3. Easier to maintain
4. Enhanced through automated reasoning
5. Because itβs not COBOL π
28.03.2025 21:51
π 1
π 0
π¬ 0
π 0
ML predicts from data, while Machine Understanding (MU) enables machines to reason and provide explainable results. MU relies on semantic & linked data, offering trust, transparency, and precision where it's most needed.
#MachineLearning #MachineUnderstanding #AI #LinkedData #SemanticWeb
18.02.2025 20:46
π 0
π 0
π¬ 0
π 0
In a post-factual world, fragile imposters exploit fear to seize power, thriving in a void created by the absence of honesty and bravery.
#socialimpact #bravenewworld
10.02.2025 15:00
π 0
π 0
π¬ 0
π 0
The national government's role in identity management may be overrated. Independent organizations or reputable private companies could take the lead - if permitted. Citizens should have the freedom to choose the service provider that best fits their needs. #decentralization #digitalidentity #ssi
03.02.2025 22:38
π 0
π 0
π¬ 0
π 0
The most effective way to boost the adoption of privacy-preserving, #self-sovereign digital #identity is through the democratization of the identity market. Opening the market to private companies would foster competition driving innovation to deliver products that customers value most.
25.01.2025 23:26
π 0
π 0
π¬ 0
π 0
#OpenSource is Not Free Labor. If your company benefits from open-source but chooses not to contribute back, you're not supporting the ecosystemβyouβre exploiting it. Open-source thrives on collaboration, not one-sided usage. It's time to step up and support the work that powers your success.
24.01.2025 15:10
π 0
π 0
π¬ 0
π 0
EuroStack
EuroStack
Nice, but Iβm quite skeptical about these top-down initiatives. To support the #EU competitiveness, establishing a transparent business ecosystem with mandatory requirements for using open technologies based on open standards in gvt. contracts would be far more effective. euro-stack.eu #eurostack
18.01.2025 17:18
π 0
π 0
π¬ 0
π 0
Chatbots that are supposed to assist you, can access private information and bypass protections, because you grant them the access - for example, by using a browser with an integrated #AI #chatbot while logged into your bank. This poses a significant #privacy and #security concern.
13.01.2025 17:52
π 0
π 0
π¬ 0
π 0
A general note on #JSON-LD - the lack of connection between compacted and expanded forms, with no traceability of what was expanded and into what, hinders advanced analytics and adaptive processing (+ it adds complexity).
JSON-LD repeats the same mistake as #XSLT in this regard.
#linkeddata
03.01.2025 12:29
π 0
π 0
π¬ 0
π 0
#nextjs + #firebase is so fast that, for the first time in my career, I had to add a delay to the loading indicator - otherwise, it just blinks. π
17.12.2024 00:46
π 2
π 0
π¬ 0
π 0
The day I began to appreciate AI was when I realized how effective it is at taking fragments of my ideas and turning them into a text that feels polished, sophisticated, and even as though it were written by someone more competent and intelligent than I am. #ai
12.12.2024 11:42
π 1
π 0
π¬ 0
π 0
@ngerakines.me Neat use of atproto to get a verifiable, thank you for sharing!
26.09.2024 17:00
π 1
π 0
π¬ 0
π 0
A verifiable credential signed with did:plc looks like a great way to open a whole new world to individuals. #decentralization #verifiablecredentials #digitalidentity
26.09.2024 11:31
π 0
π 0
π¬ 1
π 0
e.g. by letting a user sign a random string (generated by 3rd party app) by a click on a button and send the signed string back to the app - zero-knowledge-proof of #DID ownership.
26.09.2024 11:31
π 0
π 0
π¬ 1
π 0
I'm curious if there is a simple way to use #bsky #did plc method outside bluesky world for authentication. How to programmatically prove that did:plc:xyz belongs to the one who is presenting it? Does exist HTTP API to call without need to use #atproto?
25.09.2024 19:48
π 4
π 0
π¬ 1
π 0
Verifiable Credentials Signature Verification by apicatalog.com
How does a verifiable credentials signature verification work?
#verifiablecredentials github.com/filip26/iron...
23.09.2024 11:24
π 0
π 0
π¬ 0
π 0
We like to map things to graphs, but then we prefer to work with trees. Graphs are great to catch ideas, plans, and knowledge. When it comes to software engineering, graphs are hard to work with, graphs are too generic and add unnecessary complexity to a domain limited use-cases.
#linkeddata #rdf
02.09.2024 17:48
π 1
π 0
π¬ 0
π 0
LinkedTree Data Model 0.0.1
LTDM 0.0.1
github.com/filip26/link...
#linkeddata #rdf
26.08.2024 11:36
π 0
π 0
π¬ 0
π 0
Linked Trees
Another abstract data model to process #linkeddata. Influenced by #JSON-LD expanded and flattened document forms.
Linked tree terms and values are fully expanded but only identifiable node fragments are merged and flattened. Nodes with no identifiers are kept embedded.
Interested?
19.08.2024 16:54
π 0
π 0
π¬ 0
π 0
General vs Domain Vocabularies and Translators
www.reddit.com/r/semanticwe...
#linkeddata #semanticweb #rdf
30.06.2024 23:53
π 0
π 0
π¬ 0
π 0
Do you want to increase the number of forks of your GitHub #opensource project in a short period of time? Ask for donations!
27.06.2024 13:16
π 0
π 0
π¬ 0
π 0
LEXREX
Semantic vocabularies visual builder and manager
π΅οΈAnd Now for Something Completely Differentπ΅οΈ
LEXREX (~ Lexicon Rex) lexrex.web.app
Just another attempt to bring tooling into the world of semantic web. Feedback is very welcomed and appreciated.
04.06.2024 11:45
π 0
π 0
π¬ 0
π 0
GitHub - filip26/iron-verifiable-credentials: A Verifiable Credentials Processor & API
A Verifiable Credentials Processor & API. Contribute to filip26/iron-verifiable-credentials development by creating an account on GitHub.
Latest Iron VC allows you to verify and issue W3C verifiable credentials with ECDSA-RDFC-2019 (P-256, P-384), EdDSA-RDFC-2022, and Ed25519Signature2020 signature suites.Β A suite enabling a selective disclosure using NIST approved ECDSA curve coming soon!
github.com/filip26/iron...
16.02.2024 00:38
π 0
π 0
π¬ 0
π 0