3. DAN LIKES VANILLA SLICES
Subject
Predicate
Object
| February 2013 | LDP and ontology
4. DAN LIKES CAKES
What is a “Dan”?
Example: Dan is a Person
What does “likes” mean?
Example: People have preferences
Example: Like is a type of preference
| February 2013 | LDP and ontology
5. RECOMMENDATIONS AND
RULES
Using data to make recommendations
Example: Vanilla slices taste of vanilla custard
Describing things using rules
Dave is a problem
| February 2013 | LDP and ontology
6. INTERMISSION:
THE PROBLEM OF DAVE
Did I ever tell you that Mrs. McCave
Had twenty-three sons, and she named them all Dave?
Well, she did. And that wasn't a smart thing to do.
You see, when she wants one, and calls out "Yoo-Hoo!
Come into the house, Dave!" she doesn't get one.
All twenty-three Daves of hers come on the run!
7. NAMED RESOURCES
Did I ever tell you that Mrs. McCave
Had nine famous sons, and she named them all Dave?
Well, she did. And that wasn't a smart thing to do.
You see, when she wants one, and calls out "Yoo-Hoo!
Come into the house, Dave!" she doesn't get one.
All nine famous Daves of hers come on the run!
Dave_1 Dave_2
Dave_3
Dave_4
Dave_6
Dave_5
Dave_7
Dave_8
Dave_9
8. POINTING AT THINGS
Describing things using rules
Dave is a problem
Class Propertie
s
| February 2013 | LDP and ontology
9. THERE ARE DIFFERENT
KINDS OF STATEMENTS
Dave played for Manchester United
Dave plays for Paris St Germain
Triples take a snapshot – a slice through time.
| February 2013 | LDP and ontology
10. COPING WITH CONTEXT
Example: Resource_1 is a CV_event
Example: Resource_1 has agent Dave
Example: Resource_1 has duration 1995-2003
Example: Resource_1 has Game_1
And…
Example: Game_1 has player Ryan Giggs
Example: Game_1 has player Paul Scholes
| February 2013 | LDP and ontology
Editor's Notes
A triple is a fact.
But they’re mainly useful in combination. The semantic web is all about sharing. Sharing facts and inferences within web applications and across the web, lots of people sharing triples about the same sorts of things, with agreed rules, so that we can all benefit from a shared web of knowledge (factual statements).
Through sharing the semantic web becomes a ‘knowledge graph’. it combines triples – and it also leaves spaces for filling.
We need to make the implicit explicit. So we describe things in rules, which define what the triples actually mean. For example…
Triples can populate the spaces left by each other
Computers can also infer… but only if we have established all the rules – so that it can automatically fill in the gaps.
So if computers can make inferences – we need to be specific.
We do this with ontologies – but first a diversion.
We need to disambiguate between entities, so we create resources for each unique thing.
This is the stage we’re currently at. We’ve designed a simple ontology to point at things and describe them in the most basic terms. So, Dave is a person – not footballer or a anything else, just a person.
We’re using BBC ontologies including BBC, core concepts and we’re also beginning to build our own ontology.
Then we’ll move onto place and events – although the name of this are harder to
The problem is that once we have disambiguated between all the other Dave’s we also need to be careful about the statements we make about the specific Dave.
We may need to be able to disambiguate between the different ‘instances’ of Dave.
Named graph – it knows who is saying it and when it was said – which is some information, but not enough.
This is what we’re doing next. Creating ways to ensure that we can describe context effectively and that we’re always referring to the right Dave.