[GWH19] Visualizing Readable Instance Graphs of Ontology with Memo Graph
Conférence Internationale avec comité de lecture :
ICONIP 2019,
December 2019,
pp.12 pages,
Sydney,
Australia,
Mots clés: Ontology visualization, Alzheimer
Résumé:
In the context of the Captain Memo memory prosthesis for Al
z-
heimer’s patients, we want to
generate
the family
/
entourage tree of the
user
from data structured based on the PersonLink ontology.
This
graph ought to be
accessible and readable to this
particular user. In our previous work, we pr
o-
posed an ontology visualization tool called Memo Graph. It aims to offer an a
c-
cessible visualization to Alzheimer’s patients. In this paper,
w
e
extend
it
to a
d-
dress the readability req
uirement
based on
the
IKIE
V
approach. It
extracts the
most important instances (key
-
instances)
from ontology and generates a
“su
m-
mary instance graph”
(middle
-
out browsing method). The extraction and visua
l-
ization processes are undertaken incrementally. First, an “initial summary i
n-
stance graph” is generated, then permitting iteratively the visualization of su
p-
plementary key
-
instances as required.
Key
-
instances’
extraction
is based on
measures that take into account the semantic similarity between the ontological
elements and
the
use
r’s navigation history
.
This tool is integrated in
Captain
Memo
to visualize data structured using PersonLink. It is also applied for vis
u-
alizing a large
-
scale dataset DBpedia for ontology expert
s
and it is
integrated in
the context of an
approach for ling
uistic experts to visualize the LingOnto o
n-
tology. We discuss the results of the evaluation of
the
IKIEV approach