Ontologie als konkretisierte Darstellung der Wirklichkeit

1
Ontologie als konkretisierte Darstellung der
Wirklichkeit

• Barry Smith

2
MeSH

• Medical Subject Headings
• National Library of Medicine

3
What MeSH is for

• Indexing (Tagging) Medical Literature

4

• MeSH Descriptors
• Index Medicus Descriptor
• Anthropology, Education, Sociology and
  Social Phenomena (MeSH Category)
• Social Sciences
• Political Systems
• National Socialism

5
What (bio-)ontologies are for
6
what molecular function ?
what disease process ?
need for semantic annotation of data
7
need for semantic annotation of data
through labels (nouns, noun phrases) which are
algorithmically processable
8
(No Transcript)
9
warum ist die Gene Ontologie so erfolgreich?
10
(No Transcript)
11
natural language tags
to make the data cognitively accessible to human
beings
12
compare legends for maps
compare legends for maps
13
or legends for cartoons
14
(No Transcript)
15
(No Transcript)
16
ontologies are legends for data
17
ontologies are legends for images
18
what brain region ?
what brain function ?
19
ontologies are legends for mathematical equations
xi vector of measurements of gene i k the
state of the gene ( as on or off) ?i set of
parameters of the Gaussian model ... ...
20
The Gene Ontology as Integrator
GlyProt
MouseEcotope
sphingolipid transporter activity
DiabetInGene
GluChem
21
annotation using common ontologies yields
integration of databases
GlyProt
MouseEcotope
Holliday junction helicase complex
DiabetInGene
GluChem
22
annotation using common ontologies can yield
integration of image data
23
annotation using common ontologies can support
comparison of image data
24
truth
25
simple representations can be true
26
there are true cartoons
27
a cartoon can be a veridical representation of
reality
28
Cartographic Projection
29
maps may be correct by reflecting topology,
rather than geometry
30
an image can be a veridical representation of
reality
a fully labeled image can be an even more
veridical representation of reality
31
cartoons, like maps, always have a certain
threshold of granularity
32
grain resolution
33
grain resolution serves cognitive accessibility
we transform true imagesinto true cartoons
34
instances vs. types
35
two kinds of annotations
36
names of instances
37
names of types
38
pathway maps are representations of complexes of
types
39
molecular images and radiographic images are
representations of instances
40
MIAKT system
41
(No Transcript)
42
(No Transcript)
43
(No Transcript)
44
(No Transcript)
45
Patient 47920
46
(No Transcript)
47
Mammography 31667
48
Medical-Image 44922
Mammography 31667
49
Patient 47920
Medical-Image 44922
Breast 1388
MRI-Exam 32388
Mammography 31667
50
There is only one reality

• but many different representations thereof,
  including many different ontologies
• different ontologies can be simultaneously
  veridical, e.g. because of non-overlapping
  domains, or because of differential selection
• (multi-perspectivalism)

51
Reality exists before any representation
52
The ontologist becomes with a mental
representation (Bild)
53
a part of which he concretizes in an ontology

• both mental representation and ontology refer to
  the same reality

1
54
A second ontologist makes a different selection
55
Veridical ontologies can always be mapped by
taking reality as benchmark
56
but ontologies can be non-veridical, because of
different kinds of errors

• errors of mismatch with reality
• errors of understanding
• errors of coding

57
Typology of errors (Werner Ceusters)
58
http//org.buffalo.edu