Outline

1

Motivation

2

Transfer Learning Framework

3

Experimental Setup

4

Results

5

Conclusions and Future Work



2013-09-16

2 / 29

Outline

1

Motivation

2


3

Experimental Setup

4

Results

5




2013-09-16

3 / 29

Why Link Discovery?
1

2

3

Fourth Linked Data
principle
Links are central for
Cross-ontology QA
Data Integration
Reasoning
Federated Queries
...
2011 topology of the
LOD Cloud:
31+ billion triples
≈ 0.5 billion links
owl:sameAs in most
cases



2013-09-16

4 / 29

Why is it dicult?
Denition (Link Discovery)

S and T of resources and relation R
Find M = {(s , t ) ∈ S × T : R(s , t )}

Given sets

Task:
Common approaches:
Find
= {( , ) ∈
= {( , ) ∈
Find

M
M

s t S × T : σ(s , t ) ≥ θ}
s t S × T : δ(s , t ) ≤ θ}



2013-09-16

5 / 29

Why is it dicult?
Denition (Link Discovery)

S and T of resources and relation R
Find M = {(s , t ) ∈ S × T : R(s , t )}

Given sets

Task:
Common approaches:
Find
= {( , ) ∈
= {( , ) ∈
Find

M
M

1

s t S × T : σ(s , t ) ≥ θ}
s t S × T : δ(s , t ) ≤ θ}

Time complexity

Large number of triples
Quadratic a-priori runtime
69 days for mapping cities from
DBpedia to Geonames (1ms per
comparison)
Decades for linking DBpedia and LGD
...



2013-09-16

5 / 29

Why is it dicult?
2

Complexity of specications

Combination of several attributes required for high precision
Tedious discovery of most adequate mapping
Dataset-dependent similarity functions



2013-09-16

6 / 29

LIMES Framework



2013-09-16

7 / 29

Link Specication
Detection of accurate link specication is key
Link Specications has three components:
Two sets of restrictions RS ... RS resp. RT ... RT that specify the
m
1
1
k
sets resp. ,
A specication of a complex similarity metric σ via the combination of
several atomic similarity measures σ1 , ..., σn and
A set of thresholds τ1 , ..., τn such that τi is the threshold for σi .

S

T



2013-09-16

8 / 29

Transfer Learning
Classical Learning of Link Specs

Current Linking Task

Diﬀerent Linking Tasks

Task Repository

spec accuracy: α
class similarity: ζ
property similarity: π
Learning System

Learning System

Learning System

In our approach we use

Transfer Learning System

Transductive Transfer Learning

Class and property matching is assumed to be known already
(numerous approaches from ontology matching can be employed) the goal is to nd the complex similarity metric



2013-09-16

9 / 29

Outline

1

Motivation

2


3

Experimental Setup

4

Results

5




2013-09-16

10 / 29

Transfer Learning Framework I

Transfer Learning of link specications is reduce to three subproblems:
Restrictions/class similarity ζ : 2C × 2C → [0, 1]
e.g. ζ({City , Village }, {Town}) = 0.6
Property similarity: ξ : 2P × 2P → [0, 1]
e.g. ξ({rdfs : label }, {rdfs : label }) = 1.0

Accuracy of link specications: α : Q → [0, 1]



2013-09-16

11 / 29

Transfer Learning Framework II
Overall similarity measure for transfer learning:
ω(t , t ) = α(q ) · ζ(ψ(q ), C) · ζ(ψ (q ), C ) · ξ(sp (q ), PL ) · ξ(tp (q ), PL )
(details in paper)
Each similarity measure can be implemented in manifold approaches



2013-09-16

12 / 29

(details in paper)
Implementations of class similarity function ζ in framework:
label-based similarity
name-based similarity (URI similarity)
data-centric similarity
Properties similarities ξ are dened analogously



2013-09-16

12 / 29

(details in paper)
Similarities between single classes/properties can be extended to sets
(e.g. using arithmetic / geometric mean of max. similarity)



2013-09-16

12 / 29

(details in paper)
Similarities between single classes/properties can be extended to sets
(e.g. using arithmetic / geometric mean of max. similarity)
Spec can be transferred by replacing properties with most similar
properties in PL and PL


2013-09-16

12 / 29

Example (New Link Task)
Example link specication for mapping drugs in two datasets DBpedia and
Drugbank (DBpedia-Drugbank.xml):



2013-09-16

13 / 29

Example (Restriction part)
Three parts of link specs:
Restrictions part



2013-09-16

14 / 29

Example (Properties Part)
Restrictions part
Properties part



2013-09-16

15 / 29

Example (Similarities Measures Part)
Restrictions part
Properties part
Similarity Measures part: similarity metric and thresholds



2013-09-16

16 / 29

Example (Link Repository)
Transfer learning is applied using a repository → restrictions and relevant
properties are assumed to be known → nd the similarity measure by
comparing with all specs in the repository, e.g. DBpedia-SiderDrugs.xml



2013-09-16

17 / 29

Example (Restriction Similarities)
Restrictions in both specications les
Type

DBpedia-Drugbank.xml

DBpedia-SiderDrugs.xml

Source
Target

rdf:type dbpedia-owl:Drug
rdf:type drug:drugs

rdf:type sider:drugs

Straightforward label/URI similarity
For instance, trigram metric in URI similarity without prexes:
ζ({dbpedia-owl:Drug}, {dbpedia-owl:Drug}) = 1.0
ζ({sider:drugs}, {drug:drugs}) = 1.0



2013-09-16

18 / 29

Example (Restriction Similarities)
Restrictions in both specications les
Type



Source
Target

rdf:type drug:drugs

rdf:type sider:drugs

Straightforward label/URI similarity
For instance, trigram metric in URI similarity without prexes:
ζ({dbpedia-owl:Drug}, {dbpedia-owl:Drug}) = 1.0
ζ({sider:drugs}, {drug:drugs}) = 1.0

1
Data-centric: ζd (s , s ) = |P (s )||P (s

sim(x , y ) where
x ∈P (s ) y ∈P (s )
P (s ) = {x : s p x ∧ p rdf:type owl:DatatypeProperty}
(extends similarity to instances)


)|


2013-09-16

18 / 29

Example (Property Similarities)

type



Source

rdfs:label

Target

rdfs:label
drug:genericName

rdfs:label
foaf:name
rdfs:label

Applying similarity function to all properties:
For instance trigram based on URIs and arithmetic mean as
aggregation:
ξ({rdfs : label }, {rdfs : label , foaf : name }) = 0.9
ξ({rdfs : label , drug : genericName }, {rdfs : label }) = 0.8



2013-09-16

19 / 29

Example (Overall Similarity)

Based on, e.g. F-score assign quality value to q =
DBpedia-SiderDrugs.xml, in our case α(q ) = 0.89
The nal step is calculating the overall similarity measure
ω(DBpedia − Drugbank .xml , DBpedia − SiderDrugs .xml ) =
0.89 * 1.0 * 1.0 * 0.9 * 0.8 = 0.64
The steps are repeated for all link specications in the repository
Most similar link spec can be transferred by replacing its properties
with the most similar ones in the computed property matching



2013-09-16

20 / 29

Outline

1

Motivation

2


3

Experimental Setup

4

Results

5




2013-09-16

21 / 29

Experimental Setup I
The goal of evaluation is two-fold:
Evaluating whether transfer learning can be used to build templates
for link spec
Discover whether the transferred templates can be used directly
113 specications were retrieved from LATC, each has manual links
evaluation
15%

10%

3%

2%
Persons

1%
3%

Events
Locations
Diseases
Drugs
Organizations
Misc
66%



2013-09-16

22 / 29

Experimental Setup II

Leave-one-out evaluation
1.) Compare top-scored specication (most similar) and check
whether it uses the same combination of similarity functions assign 1
for match and 0 for no match
2.) Compute F-measure of learned link specs directly works only on
specs with both endpoints alive (only 12 out of 113)
Used URI similarity



2013-09-16

23 / 29

Outline

1

Motivation

2


3

Experimental Setup

4

Results

5




2013-09-16

24 / 29

First Experiments Set Results
Detecting right specication in 81% of all cases
In geo-spatial domain 91%
In persons domain 58%
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Ave

s
s
s
s
e
rag erson Event cation sease
P
Di
Lo


gs
ns
Dru izatio
n
rga
O


Mis

c

2013-09-16

25 / 29

Second Experiments Set Results
In the second Experiments series, source and target endpoints need to
be alive such that we can execute transferred link spec (12 out of 113)
In general low F-measures
100%

80%

60%

40%
Precision
Recall
F-Measure

20%

s

nt
ry

op
e

ne

i-c
ou

ty
si

ex
t

er

at
at

iv
gu
a-

pe

di
a-

te

nd

un
a-

at

pe
di
db

db

n

on

so

rs

er

pe

-p

d-

on

oo
og
f

-d

db

pe

di
a-

lin

ke

dg

eo
d

nt

or
t
rp
ai

at
a-

m

-r
st
ad
ev

en

ts

ee
r

so
n
-d

er
db

pe

di

ali

nk

ts
e
en

ev

ve

er

-e

-p

fo
od

og

ge
od

l3
s
p_

bl

er

ed

s

ty
ci

ar
ie

K-

ul

vU

ab

go

st

da
ta

di
a-

se

nt

ev
e

db

pe

di
pe

db
e-

im


-d

co
u

aco
n

es
-

-e
ss

xe
ra

eu
bc
r
rk

db

lp

-d
at

as

em

an

tic
w

eb
-

ur

re

se
ar

ch

er

nt
ry

0%


2013-09-16

26 / 29

Outline

1

Motivation

2


3

Experimental Setup

4

Results

5




2013-09-16

27 / 29

Summary
Conclusions:
Detecting right template in 81% of all cases
Transfer learning cannot replace the learning of thresholds in
specications
Future Work:
Combination with machine-learning approaches for link specications
(e.g., EAGLE, COALA), in particular for learning thresholds
More sophisticated class and property similarity approaches



2013-09-16

28 / 29

The End
Jens Lehmann
lehmann@informatik.uni-leipzig.de
AKSW/Uni Leipzig

Questions

Geo

Know
http://geoknow.eu



2013-09-16

29 / 29

Towards Transfer Learning of Link Specifications

More Related Content

What's hot (19)

Similar to Towards Transfer Learning of Link Specifications (20)

More from geoknow (13)

Recently uploaded (20)

Towards Transfer Learning of Link Specifications