What is a Knowledge Graph?
Вставка
- Опубліковано 29 вер 2024
- Learn more about Knowledge Graph → ibm.biz/knowled...
Watch "What is Natural Language Processing?" lightboard video → • What is NLP (Natural L...
Build a domain specific Knowledge Graph → ibm.biz/build-k...
Check out IBM Watson Discovery → ibm.biz/prod-wa...
Knowledge graphs represent a network of real-world entities, such as people, places, and things in the world, and illustrates the relationship between them.
In this lightboard video, Martin Keen with IBM visually explains the fundamentals of knowledge graphs and then demonstrates how they can turn your data into machine understandable language.
Get started on IBM Cloud at no cost → ibm.biz/BdftGB
Subscribe to see more videos like this in the future → ibm.biz/subscri...
#KnowledgeGraph #NaturalLanguageProcessing #DataIntegration
The most interesting and the innovative vibe guy in the entire series of these videos
This man knew how to write backwards but pointed to the wrong direction (subjectively) to the recommended list of videos...
PHENOMENAL!
"Apparently" he was writing forward during shooting and the video was flipped in post editting.
my brain matrix is unable to process the graphics after reading this.
@@li0near Nope. If you imagine yourself in his position, it is still writing backwards (right to left) and the letters would have to be reversed.
Not only real-world entities but also abstract concepts and commomsense knowledge 🧐
I remember my university professor describing ontology - this knowledge graphs seems to be something similar?
How does this relate to Object Orientated Programming?
Right handed
Well, that was interesting and well presented, thanks! I’ve been looking for a way of encoding semantic relationships to data topologies and this should be a very fruitful avenue of study. Thanks!
Hi.i am interested in knowledge graph.
Can anyone tell me a good course related to knowledge graph
These series are awesome.
Thanks. Can you tell whats graphRAG?
All this time i was dazzled by his easy of writing in mirror opposite and how ingenuine it was on the part of creator team to make him stand behind a glass to write.
They flip the video later
If you're interested in how knowledge graphs are used and how the multiple edges work I found this datacamp tutorial useful which goes into more detail: Using a Knowledge Graph to Implement a RAG Application
Hello Martin and IBM Technology, I am working on a problem where I am storing data into Knowledge Graph. But the Relation extraction is a big deal. I have got ~40% accuracy there. Could you please provide some resources that can help me in the journey? I have a .txt file that contains information about Kubernetes Security. Right now I am extracting entities (Nodes) using Spacy and feeding them to the relation extraction model to get the relations (Edges) among the Nodes. Any help would be appreciated.
did you solve the issue?
Sure, but how do you construct one? Especially not manually. And how do you do searches, and know when your done searching the graph???
I want to change nodes to objects and objects info methods and values (Json)...and then have ai present its views as characters with historic relations....streets (including transports) as roads and topics as buildings and connections as transport as meta-be-ings with their portal maybe with a trust assistant with you from birth...a guardian ai angle.
nice coffee tipp at the end :)
One of the many projects I would like IBM do for society is to make it's UML (Rational Rose)...tool intoi a SAAS with minimal cost and high availability....with interactive real time instance dashbords and present it with fun graphics and sounds....and voice commands...maybe in the MetaVerseAi bubbles.
I'm curious about injesting standards from xml, rdf, and owl and treating standards on a priority basis (RAG ?)
Thanks for the helpful video! I was kinda surprised to find you outside the brewing world :D
very clear and easy to follow - Will be viewing many more videos from this source
We use to call them semantic networks.
As long as the complete and correct meaning is conveyed.
So 2 nodes can have multiple edges?
You & coffee - not only consume, but also grow; sell; purchase; offer; make; etc.?
Paris - Roman Empire instead of Europe 🤦♂
I believe in having the edge feature endpoint objects allowing for constraints being visualized similar to (Texas Instruments stds) and that the graphic representation of all objects allow for color,,width, and height to be programmatically altered.
I like the concept of treating edges the same as nodes both expressed as UML objects which have properties and methods....,I also want their representation to reflect real time properties and constraints expressed similar to UML but when triggered flags are raised whch prompts for response or merely logs the event.
Why can't knowledge graph remember information about me?
So if the model can detect a fraudulent insurance claim, does that mean that the CIA and other secret agencies are going to have to generate a lot more detail to their cover identities?
is it formalized the same way like word embeddings? or is it a computer science concept, like graph data structure?
Ive never imagined that I will learn something dlfrom a floatinf body
How would I make use of "open linked data" or RDF files to build a knowledge graph? Specifically the Art & Architecture Thesaurus (Getty Research Institute)
I ❤️ IBM education videos
coffe is the only foundation after 5 pm :)
Well explained, it was good preparation before my first workshop about knowledge graphs tomorrow
The human brain is better.
I want to gamify weighted , directional graphs with real time animations.
Graphs should be taught in high school and the latest and greatest software and equipment made available to schools and libraries.
used with ai it could allow easy explain and learn from modifications.
presenting ai as a character that gives suggestions of paths...
"Oh Clippy meet Siri and meet the other bots and data corpuses..."
Could please tell what's this course or UA-cam playlist of this guy? Soooo good👌
Thank you. Very nice!
Awesome channel
would love to hear about CNN as well, thanks
Luckily enough, we have you covered there! We hope this helps. ua-cam.com/video/QzY57FaENXg/v-deo.html
@@IBMTechnology Nice, thanks for the link! Learnt something new.