Visualizations using Visualbox

VISUALIZATIONS WITH
SPARQL AND
VISUALBOX
Alvaro Graves
@alvarograves
gravea@rpi.edu

MOTIVATION
Linked Data brings tons of multidimensional data that
is easy(*) to query
However, having to make use of all that data we still
need to being able to process it (as humans)
One good way to make sense of this data is by using
visualizations

(*) Once you learn SPARQL

EXAMPLE

versus

Images: Few, Stephen (2010): Data Visualization for Human Perception. In: Soegaard, Mads and Dam, Rikke Friis (eds.). "Encyclopedia of Human-Computer Interaction". Aarhus, Denmark:
The Interaction Design Foundation. Available online at http://www.interaction-design.org/encyclopedia/data_visualization_for_human_perception.html

VISUALBOX
Based on LODSPeaKr
Make use of new and already existing visualization
filters
Creating of visualization via new GUI
Principle: Good data representation leads to good/easy
data manipulation
Corolary: Effort should be focused on obtaining the
right data

MODELS (SPARQL QUERY)
Using the SELECT query form will return a table, similar
to SQL.
main.query

PREFIX foaf: <http://xmlns.com/foaf/0.1/>

SELECT ?person1 ?person2 WHERE{
?person1 foaf:knows ?person2 .
}

person1 person2
http://example.org/john http://example.org/paul
http://example.org/john http://example.org/ringo
http://example.org/george http://example.org/paul

VIEWS (TEMPLATES)
We can decide how to operate with the data.
html.template

<ul>
{{for row in models.main}}
<li>{{row.person1.value}} knows {{row.person2.value}}</li>
{{endfor}}
</ul>

output

http://example.org/john knows http://example.org/paul

http://example.org/john knows http://example.org/ringo

http://example.org/george knows http://example.org/paul

FILTERS
It is possible to apply filters to the data.
Template

{{for row in models.main}}
{{row.person1.value|upper}}
{{endfor}}

Output

HTTP://EXAMPLE.ORG/JOHN
HTTP://EXAMPLE.ORG/JOHN
HTTP://EXAMPLE.ORG/GEORGE

VISUALIZATION FILTERS
It is possible to apply visualization filters directly to all
the results. These filters will generate the necessary
code to create a visualization
Model (main.query)

PREFIX cat: <http://dbpedia.org/resource/Category:>

SELECT ?countryLabel (COUNT(?nobel) as ?total) WHERE {
?nobel dcterms:subject cat:Nobel_laureates_in_Physics;
a foaf:Person;
dbp:placeOfBirth ?country .

?country a schema:Country ;
rdfs:label ?countryLabel
FILTER(LANG(?countryLabel) = "en")
}GROUP BY ?country ?countryLabel
ORDER BY DESC(?total)
LIMIT 100

countryLabel total
United States 23
Germany 14
England 9
Japan 6
Austria-Hungary 5
France 5
Netherlands 4
German Empire 3
Soviet Union 3
United Kingdom 3
Italy 3

View (html.template)

<body>
<h2>Total of Nobel laureates in Physics by country</h2>
{{models.main|GoogleVizPieChart:"countryLabel,total"}}
</body>

TABULAR DATA
Easiest case: consider table-based visualizations.

{{models.main|GoogleVizColumnChart:"countryLabel,total"}}

MAPS
Represent data as latitude, longitude and label

SELECT DISTINCT ?city SAMPLE(?lat) AS ?latitude SAMPLE(?long) AS ?longitude ?area WHERE{
?city a sch:City ;
<http://dbpedia.org/ontology/country> <http://dbpedia.org/resource/United_States> ;
geo:lat ?lat ;
geo:long ?long;
dbp:areaTotalKm ?area .

}GROUP BY ?city ?area
ORDER BY DESC(?area)
LIMIT 10

latitude longitude area
45.4983 -91.7389 25.0
30.3167 -81.6667 2292
45.9989 -112.53 1868

PARALLEL COORDINATES
Query

SELECT ?carLabel ?wheelbase ?carWidth ?carLength ?transmission WHERE {
?car dcterms:subject <http://dbpedia.org/resource/Category:Lamborghini_vehicles>;
<http://dbpedia.org/ontology/Automobile/wheelbase> ?wheelbase;
<http://dbpedia.org/ontology/MeanOfTransportation/height> ?carWidth;
<http://dbpedia.org/ontology/MeanOfTransportation/length> ?carLength;
<http://dbpedia.org/property/transmission> ?transmission;
rdfs:label ?carLabel.
FILTER(LANG(?carLabel) = "en")
}

PARALLEL COORDINATES
carLabel wheelbase carWidth carLength transmission
Lamborghini Aventador 2700.0 1136.0 4780.0 7
Lamborghini Reventón 2665.0 1135.0 4700.0 6
Lamborghini 400GT 2550.0 1257.0 4470.0 5
Template

{{models.main|D3ParallelCoordinates:"carLabel,wheelbase,carWidth,carLength,transmission"}}

GRAPHS
Question: How to express a graph in a table?
Answer: Table child → parent
Model (main.query)

SELECT ?person1 ?person2 WHERE{
?person1 foaf:knows ?person2 .
}

person1 person2
Elizabeth Engstrom Ray Bradbury
Neil Gaiman Robert A. Heinlein
Neil Gaiman Ray Bradbury

TREE STRUCTURES
Problem: How do we retrieve a tree as a table?
Solution (so far): Table child,parent with one row with no
parent (root)
child parent area

Averill Park, New York Rensselaer County, New York 8.02896e+06

Sand Lake, New York Rensselaer County, New York 9.36e+07

Schaghticoke (town), New York Rensselaer County, New York 1.3442e+08

Poestenkill (town), New York Rensselaer County, New York 8.44336e+07

Schaghticoke (village), New York Rensselaer County, New York 2.33099e+06

Rensselaer County, New York

TREE STRUCTURES
Query

PREFIX d: <http://dbpedia.org/ontology/>
SELECT max(?regionLabel) as ?child max(?superregionLabel) as ?parent (max(?totalArea) as ?area) WHERE{
{
?region d:isPartOf <http://dbpedia.org/resource/Rensselaer_County,_New_York>;
d:areaTotal ?totalArea ;
rdfs:label ?regionLabel;
d:isPartOf ?superregion .
?superregion rdfs:label ?superregionLabel.
FILTER(?superregion = <http://dbpedia.org/resource/Rensselaer_County,_New_York>)
FILTER(lang(?superregionLabel) = "en")
}UNION{
?region rdfs:label ?regionLabel .
FILTER(?region = <http://dbpedia.org/resource/Rensselaer_County,_New_York>)
}
FILTER(lang(?regionLabel) = "en")
}GROUP BY ?regionLabel ?superregionLabel

Template

{{models.main|D3CirclePacking:"child,parent,area"}}

Visualizations using Visualbox

TREE STRUCTURES (2)
Query (neruda.query)

PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX dbp: <http://dbpedia.org/ontology/>
PREFIX d: <http://dbpedia.org/resource/>
SELECT DISTINCT ?child ?parent WHERE{
{
?childNode dbp:influencedBy ?mid ;
rdfs:label ?child .
?mid dbp:influencedBy ?parentNode;
rdfs:label ?parent .
FILTER(LANG(?parent) = "en" && ?parentNode = d:Pablo_Neruda)
}UNION{
?childNode dbp:influencedBy d:Pablo_Neruda;
dbp:influencedBy ?parentNode;
rdfs:label ?child .
?parentNode rdfs:label ?parent .
FILTER(LANG(?parent) = "en" && ?parentNode = d:Pablo_Neruda)
}UNION{
d:Pablo_Neruda rdfs:label ?child
}
FILTER(LANG(?child) = "en")
}

child parent

Pablo Neruda

Dane Zajc Pablo Neruda

Gary Soto Pablo Neruda

James Tate (writer) Pablo Neruda

Richard Aitson Pablo Neruda

Erin Siegal Pablo Neruda

Jože Snoj Dane Zajc

Rudi Šeligo Dane Zajc

Veno Taufer Dane Zajc

Rigoberto González Gary Soto

Thomas Lux James Tate (writer)

html.template

{{models.main|D3Dendrogram:"child,parent"}}

NOW YOU CREATE YOUR OWN
VISUALIZATION
Use data that is of interest for you
Describe to tell a story or support a statement
Give me feedback on how does visualbox works for
you

Visualizations using Visualbox

Recommended

More Related Content

Viewers also liked (8)

Similar to Visualizations using Visualbox (20)

More from Alvaro Graves (11)

Recently uploaded (20)

Visualizations using Visualbox