A Conceptual Model for a Web-scale Entity Name System - PowerPoint PPT Presentation

1 / 26
About This Presentation
Title:

A Conceptual Model for a Web-scale Entity Name System

Description:

A Conceptual Model for a Web-scale Entity Name System Paolo Bouquet, Themis Palpanas, Heiko Stoermer, Massimiliano Vignolo University of Trento – PowerPoint PPT presentation

Number of Views:304
Avg rating:3.0/5.0
Slides: 27
Provided by: Massimilia94
Category:

less

Transcript and Presenter's Notes

Title: A Conceptual Model for a Web-scale Entity Name System


1
A Conceptual Model for a Web-scale Entity Name
System Paolo Bouquet, Themis Palpanas, Heiko
Stoermer, Massimiliano Vignolo University of
Trento bouquet, themis, stoermer,
vignolo_at_disi.unitn.it
2
Naming on the web
  • Naming by access
  • Naming by description
  • Naming by direct reference

3
Semantic webURIs identifiers for resources
  • Web resources
  • http//dbpedia.org/page/Vienna
  • HTML page
  • http//dbpedia.org/data/Vienna
  • RDF dataset
  • Non-Web resources
  • http//dbpedia.org/resource/Vienna
  • The city, the capital of Austria

4
Naming web resourcesNaming by access
Web resource
URI
accesses
http//dbpedia.org/data/Vienna
http//dbpedia.org/data/Vienna
The URI provides a causal pathway to the named
thing, mediated by the Web
5
Naming non-web resourcesNaming by description
The URI for a non-web resource redirects to a URI
naming (by access) a web resource that gives a
description of the non-web resource.
6
Semantic URI and naming by description
Semantic URI
URI
redirect to
http//dbpedia.org/data/Vienna

Naming by description
Naming by access
http//dbpedia.org/data/Vienna
Vienna
describe
RDF dataset
Non-web resource
7
QUESTION Does the Semantic Web need a third
NAMING relation? CLAIM It does.
8
ARGUMENT FOR THE CLAIM
SCENARIO 1) Let _at_ be a vocabulary for describing
accademic positions _at_affiliation _at_teaching 2)
Suppose John works at the university UNI1 and at
the university UNI2. 3) Let be the dataset
describing John's profile at UNI1. 4) Let be
the dataset describing John's profile at UNI2.
9
Two datasets
http//UNI1/data/John
http//UNI2/data/John
Philosophy of language
UNI1
UNI2
Logic
_at_affiliation
_at_teaching
_at_affiliation
_at_teaching
http//UNI1/resource/John
http//UNI2/resource/John
10
Linking data with owlsameAS
http//UNI1/data/John
http//UNI2/data/John
http//UNI2/resource/John
http//UNI1/resource/John
owlsameAs
owlsameAs
http//UNI1/resource/John
http//UNI2/resource/John
11
Merge of the two datasets
?
http//UNI1/data/John
http//UNI2/data/John
Philosophy of language
UNI1
UNI2
Logic
_at_affiliation
_at_teaching
_at_affiliation
_at_teaching
http//UNI1/resource/John
http//UNI2/resource/John
12
Query Where does http//UNI1/resource/John
teach and what? (ltx, ygt http//UNI1/resource/J
ohn _at_affiliation x ?
http//UNI1/resource/John _at_teaching y)
13
Results
ltUniversity1, Philosophy of Languagegt
ltUniversity2, Philosophy of Languagegt
ltUniversity2, Logicgt
ltUniversity1, Logicgt
The wrong results are entailed from the semantics
of owlsameAs
14
  • Solution
  • Use Entity Name System (ENS) for linking datasets
  • Global ENS identifiers for non-web resources
  • The property enscorefer

15
ENS identifiers and enscorefer
http//UNI1/data/John
http//UNI2/data/John
ensresource/John
endresource/John
enscorefer
enscorefer
http//UNI1/resource/John
http//UNI2/resource/John
16
merge
http//UNI1/data/John
http//UNI2/data/John
Philosophy of language
Logic
UNI2
UNI1
_at_affiliation
_at_teaching
_at_teaching
_at_affiliation
http//UNI1/resource/John
http//UNI2/resource/John
enscorefer
enscorefer
ensresource/John
ensdata/John
17
Collecting information about John
1. Query (xx enscorefers
ensresource/John) 2. Results , 3. Merge the
datasets ? 4. Make queries on the merge of the
datasets Who teaches what and where? (ltx,
y, zgt x _at_affiliation y ? x _at_teaching z)
18
Results http//UNI1/resource/John,
University1, Philosophy of Language http//UNI2/
resource/John, University2, Logic
19
CLAIM ENS identifiers name non-web resources by
direct reference.
20
Naming by direct reference ? naming by
description? Architecture of the Web
Pragmatic of the
Web NO

YES
21
ENS identifiers and naming by direct reference
ENS identifier
URI
redirect to
ensresource/John
ensdata/John
Naming by description
Naming by access
Naming by direct reference
John's profile ENS repository
John
describe
?
RDF Dataset
Non-web resource
22
ENS identifiers and naming by direct reference
ENS identifier
URI
redirect to
ensdata/John
ensresource/John
Naming by access
Naming by direct reference
ENS John's profile
John
fix the reference
RDF Dataset
Non-web resource
23
Fixing the reference
The man drinking martini is handsome
The man drinking martini is a computer scientist
24
Fixing the reference
http//UNI1/data/John
http//UNI2/data/John
linked
http//UNI1/resource/John
http//UNI2/resource/John
describe the same entity
John
enscorefer
enscorefer
ensresource/John
ensdata/John
25
  • Pragmatic role of ENS identifiers
  • Use ensresource/John and the enscorefer
    property
  • for linking different datasets about John.
  • Use the query
  • (x x enscorefer ensresource/John)
  • to search all datasets about John
  • and
  • to make the merge of those datasets.

26
Thank you
Write a Comment
User Comments (0)
About PowerShow.com