On 4/14/11 6:10 AM, Marco Fossati wrote:
Kingsley,
As I want to extract literals containing full names of persons from
the LOD Cloud cache, I tried to count them with the following query:
PREFIX foaf: <http://xmlns.com/foaf/0.1/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT COUNT (DISTINCT ?pers) AS ?person COUNT (DISTINCT ?name) AS
?foaf COUNT (DISTINCT ?label) AS ?rdfs WHERE { ?pers a foaf:Person ;
foaf:name ?name ; rdfs:label ?label .}
Strangely, the count of foaf:Person instances is different from the
one you mentioned: 2,828,451 (see http://bit.ly/eBzdjh).
Do you know why? Maybe timeout reasons?
No timeouts for these kinds of queries :-)
Use:
PREFIX foaf: <http://xmlns.com/foaf/0.1/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT COUNT (DISTINCT ?pers) AS ?person COUNT (DISTINCT ?name) AS ?foaf
COUNT (DISTINCT ?label) AS ?rdfs
WHERE { ?pers a foaf:Person. optional {?pers foaf:name ?name}. optional
{?pers rdfs:label ?label} }
Remember: the data has issues re. consistent use of properties and the
existence of property values. This (orthogonally) is an issue that
ultimately sheds some light on what we've had to do re. automatically
replacing URIs with Labels in our various linked data browser pages.
Basically, optionals + reasoning at massive scales i.e., users don't
experience the complexity associated with fulfilling Linked Data UI
aesthetics which boils down to taking URIs out of sight without losing
their data conduction might.
Kingsley
Cheers,
Marco
FBK Web of Data unit
http://fbk.eu
http://wed.fbk.eu/
On 4/13/11 2:37 PM, Kingsley Idehen wrote:
On 4/13/11 4:15 AM, Bernard Vatant wrote:
Hello all
Just trying to figure what is the size of personal information
available as LOD vs billions of person profiles stored by Google,
Amazon, Facebook, LinkedIn, unameit ... in proprietary formats.
Any hint of the proportion of "living" people vs historical
characters is also welcome.
Any idea?
Bernard
--
Bernard Vatant
Senior Consultant
Vocabulary & Data Integration
Tel: +33 (0) 971 488 459
Mail: bernard.vat...@mondeca.com <mailto:bernard.vat...@mondeca.com>
----------------------------------------------------
Mondeca
3, cité Nollez 75018 Paris France
Web: http://www.mondeca.com
Blog: http://mondeca.wordpress.com
----------------------------------------------------
Bernard,
LOD Cloud cache has 3,321,094 foaf:Person entities [1]. Distinct
count 3,319,862 count [2].
URIBurner has 4,564,981 foaf:Person entities [3]. Distinct count is
4,555,697 [4] .
Both cases via SPARQL aggregate queries against their respective
endpoints. Note, no inference context applied there are a variety of
rules across OpenCyc, UMBEL, Yago, and DBpedia that would alter these
counts.
Tip re. URLs below, simply change the "authority" part of the URL
when seeking similar counts from other Virtuoso instances, with some
luck it could apply to other SPARQL endpoints in general, subject to
what the endpoints support and permit etc..
SPARQL queries used across each endpoint:
select count(?s) where {?s a foaf:Person}
select count(distinct ?s) where {?s a foaf:Person}
Links:
1. http://lod.openlinksw.com/c/CYIZZL4 -- LOD Cloud Cache
2. http://lod.openlinksw.com/c/COXER7C -- LOD Cloud Cache Distinct Count
3. http://uriburner.com/c/DYVU7N -- URIBurner
4. http://uriburner.com/c/DV6VPQ -- URIBurner Distinct Count .
--
Regards,
Kingsley Idehen
President& CEO
OpenLink Software
Web:http://www.openlinksw.com
Weblog:http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen
--
Regards,
Kingsley Idehen
President& CEO
OpenLink Software
Web: http://www.openlinksw.com
Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen