程序代写代做 algorithm graph html Semantic Technologies and Applications COMP5860M

Semantic Technologies and Applications COMP5860M
John Stell
Room 9.15, School of Computing
j.g.stell@leeds.ac.uk
Lecture 13: March 2020
“Ontology is Overrated” and other controversies
1

Today, we look at some controversies around semantics, using these sources:
􏰀 Fritz Lehmann: Semantic Networks (1992)
􏰀 Clay Shirky: Ontology is Overrated (c2005)
􏰀 Lindsay Poirier: A Turn for the Scruffy (2017)
􏰀 Abraham Bernstein et al: A new Look at the Semantic Web (2016)
2

Lehmann 1992
3

Fattipuffs and Thinifers; in semantics there are Neats and Scruffies
4

From Lehmann 1992. The Christian Trinity as Semantic Network
5

From Lehmann 1992. Frame systems; frames have slots
6

From Lehmann 1992. IS-A Hierarchy with Instances
7

From Shirky 2005. At one time, the Libary of Congress.
Shirky relates this to Books as physical objects, which need storage space and can only be in one location. And fitting things into fixed categories . . .
8

Shirky on tagging as organizing
. . . systems that are allowing for and aggregating tags. . . . they don’t recreate the structured, hierarchical categorization so often forced onto us by our physical systems. Instead, we’re dealing with a significant break – by letting users tag URLs and then aggregating those tags, we’re going to be able to build alternate organizational systems, systems that, like the Web itself, do a better job of letting individuals create value for one another, often without realizing it.
9

Bernstein et al make the point: Practical aspects of Semantic Web are in real-world use.
For example:
􏰀 schema.org
􏰀 linked data
􏰀 usage by BBC and other media (www.bbc.co.uk/ontologies) 􏰀 Knowledge Graphs used by Google etc
􏰀 Support for RDF etc in RDBMS
􏰀 Semantics used for recommender systems
􏰀 Disease terminology from WHO
􏰀 IBM Watson using knowledge from dbpedia etc
but formal semantics is not much used
10

https://schema.org/

Avatar

Director: James Cameron (born August 16, 1954)Science fiction
Traile

microdata via tags in HTML5 schema.org provides shared vocabulary
11

Bernstein et al see (saw) the future as
􏰀 less use of logic, more existing language
􏰀 ontologies learnered or enhanced automatically
A letter responding to Bernstein et al suggests
. . . Bernstein et al. seemed to neglect a key feature of formal semantics transparency. They seem comfortable with the relaxation of logic as a conceptual framework for the Semantic Web, which is typical of modern Knowledge Graphs (such as the one Google uses). But one of the consequences of such relaxation is that part of data semantics ends up being embedded in algorithms. Not only practitioners but also common users are aware that algorithms that work on Web data are embedded in only a few monolithic, private platforms that are far from open, transparent, and auditable.
12

Poirier is not a computer scientist but a cultural anthropologist
Her perspective is understanding how neat and scruffy views have evolved and the ways in which they have impact on the Semantic Web and knowledge representation on the web today.
13

Poirier quotes Bizer:
If you ask Tim, kind of everybody is shifting. But basically there’s strong deployment on the distributed data provision side, which is an integral part of the Semantic Web. There’s strong deployment on the ontology side because basically people agree on the global ontology, which is schema.org, which is not as formal. You can’t do reasoning with it … but you do some other kind of reasoning. You do more statistical reasoning, not so much based on ontology constructs. It’s more based on the data…
14

So what about Your Coursework?
It needs to be neat rather than scruffy (this is not to say scruffy is bad, just the coursework is about modelling in a formal way in OWL)
􏰀 A lot of people have asked me similar things
􏰀 There is a bit of misunderstanding about some neat things
􏰀 I’ve seen a lot of example with Classes where you have not thought: ‘what exactly are the individuals?’
􏰀 I’ve seem a lot of links between classes in diagrams that are not is-a links but are meant to be
􏰀 For example are these the same:
􏰀 The Class of novels (e.g. in a library)
􏰀 The classification ‘novel’ as one way of classifying a book
􏰀 The label “novel” in English as the name of this classification
􏰀 the relationship linking every book which is a novel to its
classification
15

Using SKOS concepts like novel you might say it’s narrower that a concept English Novelist. dbpedia has examples of this.
it’s not wrong
but don’t confuse this with Classes in OWL or Concepts in DL
Of concepts they (SKOS) say:
“The fundamental element of the SKOS vocabulary is the concept. Concepts are the units of thought ideas, meanings, or (categories of) objects and events?which underlie many knowledge organization systems. As such, concepts exist in the mind as abstract entities which are independent of the terms used to label them.”
16