SemanVc Web MoVvaVng Example

1/31/12 A Mo+va+ng example Seman+c Web Mo+va+ng Example We start with a book... • Here’s a mo+va+ng example, adapted from a presenta+on by Ivan Herman • It introduces seman+c web concepts • and illustrates the benefits of represen+ng your data using the seman+c web techniques • And mo+vates some of the seman+c web technologies A simplified bookstore data ID
ISBN 0-00-6511409-X
ID
Author
Title
Publisher
id_xyz
The Glass Palace
id_qpr
Name
id_xyz
Ghosh, Amitav
ID
id_qpr
Homepage
http://www.amitavghosh.com
Publisher’s name
Harper Collins
Year
2000
City
London
1 1/31/12 Export data as a set of rela&ons The Glass Palace 2000 London Harper Collins Notes on exporting the data
a:title
• Rela+ons form a graph h"p://…isbn/000651409X a:year
a:city
e
ish
ubl
a:p
– Nodes refer to “real” data or some literal – We’ll defer dealing with the Graph representa+on r
• Data export doesn’t necessarily mean physical conversion of the data a:author
e
a:p_nam
–  rela+ons can be generated on-­‐the-­‐fly at query +me a:name
Ghosh, Amitav • All of the data need not be exported a:homepage
h"p://www.amitavghosh.com Same book in French…
Bookstore data (dataset “F”) A
1
2
B
ID
ISBN 2020286682
C
Titre
Le Palais des Miroirs
D
Traducteur
$A12$
Original
ISBN 0-00-6511409-X
3
4
5
6
7
ID
ISBN 0-00-6511409-X
Auteur
$A11$
8
9
10
Nom
11
Ghosh, Amitav
12
Besse, Christianne
2 1/31/12 Export data as a set of rela&ons
h"p://…isbn/000651409X Start merging your data
The Glass Palace a:title
2000 a:year
h"p://…isbn/000651409X Le palais des miroirs f:o
rig
in
al
f:auteur
London e
itr
f:t
Harper Collins she
ubli
a:p
a:city
r
a:author
e
a:p_nam
a:name
h"p://…isbn/2020386682 h"p://…isbn/000651409X a:homepage
l
ina
rig
f:o
f:traducteur
Ghosh, Amitav h"p://www.amitavghosh.com e
itr
f:t
f:auteur
f:nom
Le palais des miroirs h"p://…isbn/2020386682 f:nom
Ghosh, Amitav f:traducteur
f:nom
Besse, ChrisJanne f:nom
Ghosh, Amitav Besse, ChrisJanne Merging your data
The Glass Palace a:title
2000 a:year
London Harper Collins Merging your data
h"p://…isbn/000651409X she
ubli
a:p
a:city
Same URI!
r
a:author
e
a:p_nam
The Glass Palace a:title
2000 a:yea
r
London Harper Collins a:name
h"p://www.amitavghosh.com f:auteur
r
a:author
e
a:p_nam
f:original
f:auteur
a:homepage
l
ina
rig
f:o
Ghosh, Amitav she
ubli
a:p
a:city
a:name
h"p://…isbn/000651409X a:homepage
h"p://…isbn/000651409X Le palais des miroirs Le palais des miroirs Ghosh, Amitav h"p://www.amitavghosh.com e
itr
f:t
e
itr
f:t
h"p://…isbn/2020386682 h"p://…isbn/2020386682 f:traducteur
f:traducteu
r
f:nom
Ghosh, Amitav f:nom
Besse, ChrisJanne f:no
m
Ghosh, Amitav f:nom
Besse, ChrisJanne 3 1/31/12 However, more can be achieved… Start making queries…
• User of data “F” can now ask aout the +tle of the original • This informa+on is not in the dataset “F”… • …but can be retrieved by merging with dataset “A”! • Maybe a:author & f:auteur should be the same • But an automa+c merge doesn’t know that! • Add extra informa+on to the merged data: –  a:author same as f:auteur –  both iden+fy a “Person” –  Where Person is a term that may have already been defined, e.g.: •  A “Person” is uniquely iden+fied by a full name, a homepage, facebook page, G+ page or email address •  It can be used as a “category” for certain type of resources Use this extra knowledge
The Glass Palace a:title
2000 a:year
• User of dataset “F” can now query: h"p://…isbn/000651409X Le palais des miroirs f:original
London Harper Collins a:city
a
he
blis
:pu
r
f:
a:author
e
a:p_nam
• The informa+on is not in datasets “F” or “A”… • …but was made available by: h"p://…isbn/2020386682 f:auteur
f:traducteur
r:type
a:name
a:homepage
h"p://…foaf/Person f:nom
Besse, ChrisJanne Ghosh, Amitav h"p://www.amitavghosh.com –  “donnes-­‐moi la page d’accueil de l’auteur de l’original” •  well… “give me the home page of the original’s ‘auteur’” e
titr
r:type
f:nom
This enables richer queries –  Merging datasets “A” and datasets “F” –  Adding three simple extra statements –  Inferring the consequences 4 1/31/12 Merge with Wikipedia data Combine with different datasets • Using, e.g., the “Person”, the dataset can be combined with other sources • For example, data in Wikipedia can be extracted using dedicated tools The Glass Palace 2000 a:title
h"p://…isbn/000651409X a:year
Le palais des miroirs f:original
London Harper Collins a:city
a
he
blis
:pu
r
re
f:tit
a:author
e
a:p_nam
h"p://…isbn/2020386682 f:auteur
r:type
–  e.g., the “dbpedia” project can extract the “infobox” informa+on from Wikipedia already… f:traducteur
a:name
r:type
h"p://…foaf/Person a:homepage
f:nom
f:nom
r:type
Besse, ChrisJanne Ghosh, Amitav h"p://www.amitavghosh.com foaf:name
w:reference
h"p://dbpedia.org/../Amitav_Ghosh Merge with Wikipedia data The Glass Palace 2000 a:title
The Glass Palace h"p://…isbn/000651409X 2000 a:year
London a:city
a
he
blis
:pu
r
re
f:tit
a:author
e
a:p_nam
r:type
a:name
f:nom
a:homepage
Harper Collins h"p://…foaf/Person a
h"p://dbpedia.org/../The_Glass_Palace w:reference
a:homepage
f:nom
foaf:name
Besse, ChrisJanne h"p://www.amitavghosh.com h"p://dbpedia.org/../The_Glass_Palace w:reference
w:author_of
h"p://dbpedia.org/../Amitav_Ghosh w:born_in
w:author_of
h"p://dbpedia.org/../Kolkata h"p://dbpedia.org/../The_Hungry_Tide w:long
w:author_of
h"p://dbpedia.org/../The_Calcu"a_Chromosome r:type
w:isbn
w:author_of
h"p://dbpedia.org/../The_Hungry_Tide f:traducteur
h"p://…foaf/Person r:type
Ghosh, Amitav w:author_of
h"p://dbpedia.org/../Amitav_Ghosh h"p://…isbn/2020386682 f:auteur
f:nom
Besse, ChrisJanne Le palais des miroirs re
f:tit
a:author
e
a:p_nam
a:name
f:nom
h"p://www.amitavghosh.com r
r:type
r:type
w:isbn
foaf:name
a:city
he
blis
:pu
f:traducteur
r:type
Ghosh, Amitav h"p://…isbn/000651409X a:year
f:original
London h"p://…isbn/2020386682 f:auteur
a:title
Le palais des miroirs f:original
Harper Collins Merge with Wikipedia data w:lat
w:author_of
h"p://dbpedia.org/../The_Calcu"a_Chromosome 5 1/31/12 Is that surprising?
• It may look like it but, in fact, it should not be… • What happened via automa+c means is done every day by Web users! • What is needed is a way to let machines decide when classes, proper+es and individuals are the same or different This can be even more powerful • Add extra knowledge to the merged datasets –  e.g., a full classifica+on of various types of library data –  geographical informa+on –  etc. • This is where ontologies, extra rules, etc., come in –  ontologies/rule sets can be rela+vely simple and small, or huge, or anything in between… • Even more powerful queries can be asked as a result So where is the Semantic Web?
• The Semantic Web provides technologies to
make such integration possible!
• Hopefully you get a full picture at the end of
the tutorial…
6