Addressing Issues in Foundational Ontology Mediation

Zubeida Casmod Khan and C. Maria Keet

School of Mathematics, Statistics, and Computer Science, University of KwaZulu-Natal, Durban, South Africa

UKZN/CSIR-Meraka Centre for Artiﬁcial Intelligence Research, Durban, South Africa

Keywords:

Foundational Ontology, Ontology Mediation, Semantic Interoperability, Ontology Alignment, Ontology

Mapping, Ontology Matching, Ontology Merging.

Abstract:

An approach in achieving semantic interoperability among heterogeneous systems is to offer infrastructure to

assist with linking and integration using a foundational ontology. Due to the creation of multiple foundational

ontologies, this also means linking and integrating those ones. In order to achieve this, we have selected

the widely used foundational ontologies DOLCE, BFO, and GFO, and their related modules, on which to

perform ontology mediation (alignment, mapping, and merging). The foundational ontologies were aligned

by identifying correspondences between ontology entities using seven tools, documentation, and our manual

alignments, and comparing their effectiveness. Thereafter, based on the alignments, we created correspon-

dences in the ontology ﬁles resulting in entity mappings and merged ontologies. However, during the mapping

process, it was found that differences in foundational ontologies, such as their hierarchical structure, conﬂict-

ing axioms due to complement and disjointness, and incompatible domain and range restriction, cause logical

inconsistencies in foundational ontology alignments, thereby greatly reducing the number of mappings. We

analyse and present these logical inconsistencies with possible solutions to some of them.

1 INTRODUCTION

There has been an exponential growth in ontology de-

velopment for the Semantic Web, including a move

toward modular and networked ontologies. Founda-

tional ontologies are commonly used to facilitate se-

mantic interoperability. However, Semantic Web sys-

tem developers use their ontologies with a preferred

foundational ontology. The semantics and underlying

Ontology of each foundational ontology differs, caus-

ing a problem in semantic interoperability. Heteroge-

neous systems on the Semantic Web are restricted to

committing to a single foundational ontology in order

to promote interoperability. However, no single foun-

dational ontology is used across all systems, therewith

preventing interoperability. In order for these applica-

tions to share and process information correctly, there

is a need for foundational ontology interoperability,

so that ontology developers committing to a preferred

foundational ontology will achieve seamless linking

to other domain ontologies linked to another foun-

dational ontology. Such an infrastructure was envi-

sioned as the “WonderWeb Foundational Ontologies

Library” (WFOL) (Masolo et al., 2003), but this in-

frastructure still does not exist. The main precondi-

tions for a WFOL are content comparisons and ontol-

ogy mediation. Ontology mediation refers to identi-

fying and solving differences between heterogeneous

ontologies, in order to allow reuse and interoperabil-

ity. Its three main processes are alignment, mapping,

and merging (de Bruijn et al., 2006). There are only

few paper-based alignments of foundational ontolo-

gies, being between GFO and DOLCE (Herre, 2010)

and between DOLCE and BFO (Seyed, 2009; Temal

et al., 2010), which, however, are partial, with older

versions of the ontologies, informal, and/or aligned

but not mapped. To the best of our knowledge, no

systematic comparison of the contents of foundational

ontologies has been done, nor full alignments, let

alone providing consistent mappings.

We aim to contribute to ﬁll this gap of semantic in-

teroperability by selecting three well-known founda-

tional ontologies, DOLCE (Masolo et al., 2003), BFO

(http://www.ifomis.org/bfo) with RO (Smith et al.,

2005), and GFO (Herre, 2010) with which we per-

form a rigorous foundational ontology content com-

parison and mediation to aid in achieving founda-

tional ontology interchangeability. The alignment

process is carried out by using the manual alignment

as a gold standard and (semi-)automated alignment

with seven alignment tools to examine them on their

capabilities to align foundational ontologies. The ac-

Keet C. and Dawood Z..

Addressing Issues in Foundational Ontology Mediation.

DOI: 10.5220/0004518000050016

In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2013), pages 5-16

ISBN: 978-989-8565-81-5

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

curacy and percentage of alignments that were found

vary greatly among the tools due to their diverse

alignment algorithms, ranging from 18 to 94% and 17

to 31%, respectively. Further alignment issues appear

in the transitivity of alignments across the three foun-

dational ontologies due to absence of some entity or

conﬂicting parthood theories, whilst some may be re-

solved by asserting them as sibling classes. Mapping

the aligned entities whilst keeping a consistent ontol-

ogy reduces the feasible set from 85 alignments to 43

successful mappings due to disjointness and comple-

ment axioms elsewhere in the ontology, and due to in-

compatible domain and range axioms, which in some

cases can be solved from a logic viewpoint by assert-

ing subsumption instead. For each mediation process

(alignment, mapping and mediation), we present the

issues encountered for foundational ontology media-

tion and aim to solve these issues.

In the remainder of the paper, we provide a liter-

ature review in Section 2. A high-level content com-

parison of the foundational ontologies is described in

Section 3, which is followed by an analysis of align-

ments in Section 4, and of the mappings in Section 5.

We discuss the results in Section 6 and conclude in

Section 7.

2 LITERATURE REVIEW

There has not been much work done in comparing

content of the foundational ontologies. By this, we

mean comparing their classes, properties and rela-

tions. Some work on comparing the primitive rela-

tions of BFO (i.e., the Relation Ontology (RO)) and

DOLCE has been done (Seyed, 2009) where it is

found that the philosophies behind the foundational

ontologies affect the way the relations are modelled.

For instance, BFO is based on realist principles and

has no abstract entities while GFO is both descriptive

and realist in nature and allows abstract entities in an

ontology, and BFO’s parthood relation has part does

not consider abstract entities, while GFO has a part-

hood relation abstract has part that considers abstract

entities at a higher-level than its has part relation.

Temal et al. (Temal et al., 2010) created a BFO-

DOLCE mapping in order to integrate medical in-

formation. The classes (universals or categories) are

mapped with equivalence and subsumption relations.

Based on the older so-called SNAP and SPAN version

of BFO, they found that all BFO universals were suc-

cessfully mapped to DOLCE, but not all DOLCE enti-

ties could be mapped to BFO. These alignments were

not checked on consistency of the mappings and were

done on some First Order Logic version of the on-

tologies, where the SNAP-BFO has, e.g., Boundary,

that BFO v1.1 in OWL does not have, and DOLCE

is claimed to have Collection, which appears neither

in the principal documentation (Masolo et al., 2003)

nor in the OWLized version of DOLCE. Some of their

alignments are useful, however, which we will return

to in Section 4.

From a computational viewpoint instead of an On-

tology and modelling viewpoint, we consider several

aspects of ontology mediation and matching. Ontol-

ogy mediation (de Bruijn et al., 2006) is divided into

three operations: mapping, alignment, and merging.

To be precise in the terminology we use throughout

the paper, we provide several deﬁnitions on ontology

matching in this section, which are taken from (Eu-

zenat and Shvaiko, 2007), Chapter 3. First, there is

the actual mediation, or matching, process:

Deﬁnition 1 (Matching Process). The matching pro-

cess can be seen as a function f which, from a pair of

ontologies to match o and o

, an input alignment A, a

set of parameters p and a set of oracles and resources

r, returns an alignment A

between these ontologies:

= f (o, o

, A, p, r).

To be able to talk about an actual alignment or

mapping, the notion of “entity language” has to be

introduced, which is used to express precisely those

entities that will be matched.

Deﬁnition 2 (Entity language). Given an ontology

language L, an entity language Q

is a function from

any ontology o ⊆ L which deﬁnes the matchable enti-

ties of ontology o.

Then, a correspondence consists of a relation be-

tween two entities in different ontologies, which is

uniquely identiﬁed and has some conﬁdence value as-

signed to it.

Deﬁnition 3 (Correspondence). Given two ontolo-

gies o and o

with associated entity languages Q

and Q

, a set of alignment relations θ and a conﬁ-

dence structure over Ξ, a correspondence is a 5-tuple:

hid, e, e

, r, ni, such that id is a unique identiﬁer of the

given correspondence, e ⊆ Q

(o) and e

⊆ Q

r ⊆ θ, and n ⊆ Ξ.

Ontology alignment, then, is the process of spec-

ifying correspondences between entities, by using a

particular alignment relation, such as equivalence,

subsumption, or a predeﬁned similarity relation.

Deﬁnition 4 (Alignment). Given two ontologies o

and o

, an alignment is made up of a set of correspon-

dences between pairs of entities belonging to Q

(o)

and Q

) respectively.

Ontology mapping deals with creating correspon-

dences between ontologies based on alignments such

KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

that the resultant ontology is still consistent and does

not have unsatisﬁable classes or relations. Euzenat

and Shvaiko do consider this with respect to models

of aligned ontologies, which is too lengthy to repeat

here, and De Bruijn et al. does not provide a deﬁnition

of their idea of mapping as a ‘consistent alignment in

the context of the whole ontology’ either. Therefore,

we capture the gist in the following deﬁnition, using

Euzenat and Shvaiko’s notational conventions.

Deﬁnition 5 (Mapping). Given two ontologies o and

, a mapping is made up of a set of correspon-

dences between pairs of entities belonging to Q

(o)

and Q

), respectively, and this mapping is satis-

ﬁable and does not lead to an unsatisﬁable entity in

either o or o

In merging, a new merged ontology is created

from the original ontologies.

Deﬁnition 6 (Merging). Given two ontologies o and

, a merging is the creation of a new ontology o

containing o and o

and all mappings between entities

belonging to Q

(o) and Q

) such that o

does not

have unsatisﬁable entities and is consistent.

Overviews of approaches, frameworks, and tech-

nologies used to perform ontology mapping, align-

ment and merging are discussed elsewhere (e.g.,

(de Bruijn et al., 2006)), and more detail about al-

gorithms and issues can be found in (Euzenat and

Shvaiko, 2007).

A number of tasks for the problem at hand are

based on ontology mediation. Performing founda-

tional ontology mediation with automated tools is a

good approach and a starting point because there are

many foundational ontology modules, and founda-

tional ontologies are constantly being updated. It is

rather time-consuming to explore each foundational

ontology, especially when there are differences in

hierarchy and structure. By applying several tools

to align foundational ontologies, one can determine

which tools are better suited for foundational ontolo-

gies, and the type of alignments that are misaligned

or not discovered by tools. We summarize the align-

ment tools that are used in the experimental evalua-

tion, of which we note that LogMap, YAM++, Hot-

Match, Hertuda and Optima have been evaluated with

positive results by the Ontology Alignment Evalua-

tion Initiative (OAEI) in terms of their precision, re-

call and other performance measures.

H-Match (Castano et al., 2003) is an algorithm

for matching ontologies at different depth levels, with

different accuracies, based on user choices. The algo-

rithm takes into account linguistic and semantic fea-

tures of ontologies to perform matching and uses one

of four matching models: surface, shallow, deep or in-

tensive. The surface model only considers linguistic

afﬁnity between entity names to measure similarity.

In shallow, deep and intensive models, context is also

considered to determine entity similarity.

PROMPT (Noy and Musen, 2000) is an ontology

matching plug-in for Prot

e that allows for compar-

ison, mappings, and merging between ontologies. It

is a semi-automatic method that invokes algorithms

based on a combination of concept-representation

structure, the relations between entities and user’s ac-

tions. PROMPT offers the user four different algo-

rithms to use for initial comparison: lexical match-

ing, FOAM plugin, lexical matching with synonyms

and using UMLS concept identiﬁers for matching. It

is only supported in older versions of Prot

e, which

makes it unstable.

LogMap (Jim

enez-Ruiz and Cuenca Grau, 2011)

automatically generates mappings between ontolo-

gies using logic-based semantics of the input ontolo-

gies. It offers an improvement to other mapping tools

in that it addresses scalability and logical inconsisten-

cies. LogMap allows a user to upload ontologies in a

number of formats and implements existing reasoners

to check the satisﬁability of the ontologies.

YAM++ (Ngo and Bellahsene, 2012) aligns en-

tities by information retrieval or machine learning if

training data is available. Three matchers are imple-

mented in YAM++: an element level matcher, a struc-

tural matcher and a semantic matcher. The element

level and structural mapper discover alignments while

the semantic matcher revises these alignments to re-

move inconsistencies and ensure logical mappings.

HotMatch (Dang et al., 2012) is a tool based on a

combination of many matching algorithms. The two

types of algorithms are element level and structural

matching. However, there is more than one of each

implemented. There are also ﬁlters in HotMatch, used

to remove duplicate mappings found by the matchers.

Upon input of a source and target ontology, HotMatch

deploys its matchers and ﬁlters sequentially resulting

in mappings between the two.

Hertuda (Hertling, 2012) is an entity matcher that

applies element level matching with a string compar-

ison. The alignments generated by Hertuda are only

satisﬁable in OWL Lite/DL. As a result, object prop-

erties in the ontologies are handled separately. This

may cause some difﬁculties in aligning object prop-

erties in the foundational ontologies because their do-

mains and ranges affect the alignments.

Optima (Kolli and Doshi, 2008) is a fully auto-

matic tool which iteratively improves alignments. It

is aimed at aligning large ontologies but may also be

used for smaller ontologies. Its similarity measure is

based on both syntactic and semantic similarity.

AddressingIssuesinFoundationalOntologyMediation

3 FOUNDATIONAL ONTOLOGY

CONTENT COMPARISON

In this section, we provide an informal content com-

parison between the foundational ontology pairs by

identifying differences and similarities between the

them. A content comparison is beneﬁcial in that it

forms the basis for performing ontology mediation

operations. A content comparison does not include

abstract comparisons such as those based on philo-

sophical choices, ontological alignments and software

engineering properties, which has been addressed

elsewhere (Khan and Keet, 2012), but rather a high-

level comparison of the structure, organisation, and

entities of the foundational ontologies.

DOLCE, BFO and GFO contain both 3D and 4D

entities. Both BFO and GFO name these entities Con-

tinuant and Occurrent while DOLCE names them en-

durant and perdurant. Some syntactic variants ex-

ist between DOLCE, BFO, and GFO, e.g., DOLCE’s

space-region vs. BFO’s SpatialRegion vs. GFO’s

Spatial region. In DOLCE, BFO, and GFO, classes

that share the same name and idea are process, func-

tion and role.

DOLCE entities are of type particular, BFO’s en-

tities are Universals while GFO contains a combina-

tion of the two, both Individual and Universal entities.

DOLCE and BFO have similar structures at a high-

level only in that both have separate branches of 3D

and 4D entities. GFO’s high-level structure is differ-

ent as it offers a distinction between Category and In-

dividual entities. DOLCE’s endurant and perdurant

branches are linked by participation relations; BFO’s

and GFO’s 3D and 4D entity branches are completely

independent of each other.

DOLCE and GFO have advanced support for rep-

resenting entity properties (e.g., colour) and their val-

ues (e.g., blue) while BFO has limited support for

this. To describe these entities, DOLCE uses the

terms quality, quale and quality-space, BFO uses the

term Quality while GFO names these entities Prop-

erty, Property value and Value space. DOLCE’s

quality branch is disjoint to its endurant and perdu-

rant branches. In BFO, on the other hand, Qual-

ity is subsumed by Continuant branch, while GFO

has Property subsumed by the higher-level Individ-

ual. BFO’s temporal entities, including temporal re-

gions, intervals, and instants, are subsumed by Oc-

current, while DOLCE’s temporal entities are split up

into three parts, being temporal regions that are sub-

sumed by abstract entities, temporal qualities that are

subsumed by quality entities, and subclasses of perdu-

rants. All of GFO’s temporal entities are subsumed

by its Space-time entity, which is disjoint to its Con-

crete and Abstract entities. Most of DOLCE’s spatial

entities are subsumed by its abstract entity with the

exception of spatial-location-q which is subsumed by

quality, while GFO’s spatial entities are subsumed by

its space-time entity. BFO’s SpatialRegion entity is

subsumed by Continuant. DOLCE and GFO have ab-

stract entities while BFO does not.

DOLCE and GFO contain relational properties.

BFO does not have relational properties included in

the ontology, but rather as a separate ontology, the

Relational Ontology (RO) (Smith et al., 2005). BFO

2.0 is currently being developed, where BFO is in-

tegrated with RO. DOLCE’s relational properties are

all based on either of its six primitive relations: part-

hood, temporary parthood, constitution, participation,

quality, and quale. For mereology, DOLCE adopts the

axioms of General Extensional Mereology (GEM).

BFO core is a comprehensive mereology represented

in ﬁrst-order logic and contains collections, sums and

universal axioms. GFO’s mereology contains the fol-

lowing axioms: antisymmetry, transitivity, set inclu-

sion, proper parthood, and other GFO-speciﬁc axioms

based on these.

Thus, the organisation of entities within the three

ontologies differ. In some cases, entities that seem

similar fall in contradicting or disjoint classes. These

differences in structure and organisation may cause

inconsistencies when performing mapping, as we

shall see later in detail.

4 ALIGNMENT

For foundational ontology alignment, i.e., aligning on

an entity-by-entity basis, certain aspects of the un-

derlying philosophies of each foundational ontology

have been ignored, because else it would result in few

or no alignments and for practical usage of their OWL

ﬁles, they are less pressing issues (e.g., DOLCE is

descriptive and contains particulars, while BFO is re-

alist and contains universals). We align classes and

object properties with equivalence relations ﬁrst, and

use subsumption relations afterward to resolve some

mapping inconsistencies.

We create alignments for 20 pairs of ontologies.

These ontologies include DOLCE-Lite, BFO, GFO,

FunctionalParticipation, SpatialRelations, and Tem-

poralRelations (which are more-detailed modules of

DOLCE), BFORO and GFO-Basic. BFORO refers

to the merged ontology of BFO with the RO, and

GFO-Basic is a less-detailed module of GFO. We per-

form ontology alignment by using existing tools, doc-

umentation and manually using the content compari-

son, entity axioms and annotations. Further, for each

KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

resource (tool, documentation or manual alignment),

we measure its accuracy by ﬁrstly examining each of

its output alignments to determine whether or not the

equivalence relation is correct. Thereafter we deﬁne

accuracy as the number of ‘correct’ alignments over

the total alignments given by the resource (Eq. 1),

where ‘correct’ denotes the alignment is also in the

set of alignments found manually, i.e., what is typi-

cally considered as the ‘gold standard’. We deﬁne the

found measure of the resources as the number of cor-

rect alignments over the total possible correct align-

ments, after manual intervention (Eq. 2).

Accuracy =

|correct alignments|

|total alignments

resource

× 100 (1)

Found =

|correct alignments|

|total alignments

gold

× 100 (2)

4.1 Alignment Results

We describe the results of the manual alignments ﬁrst,

and then the results obtained with the matching tools.

4.1.1 Manual Alignments

The yield of the manual alignments between the

main foundational ontologies (DOLCE-Lite, BFO

and GFO) resulted in 17 alignments for DOLCE-Lite

↔ BFO, 23 alignments for BFO ↔ GFO and 35

alignments for GFO ↔ DOLCE-Lite, hence, 75 in to-

tal; the complete set of alignments can be accessed

at http://www.thezﬁles.co.za/ROMULUS/. Of these

75 alignments, 37 are for alignments between object

properties and are displayed in Tables 1, 2, and 3. The

mapped class alignments are available and discussed

elsewhere (Khan and Keet, 2013) and online accessi-

ble, and the difference between class alignments and

mappings for GFO ↔ DOLCE-Lite are displayed in

Table 4. When we consider entity alignments includ-

ing the related modules of the foundational ontologies

(e.g., GFO-Basic), there is a total of 85 alignments.

Naturally, there are many more than 85 alignments

if we consider identical alignments that occur among

the same entities in related modules; e.g., DOLCE-

Lite:particular ↔ GFO:Individual and FunctionalPar-

ticipation:particular ↔ GFO:Individual. There are 14

alignments common between these three ontologies,

which is displayed in Table 5.

The manual alignments were aided by the GFO

documentation (Herre, 2010) and checked against the

alignments proposed by (Temal et al., 2010; Seyed,

2009). The GFO documentation (Herre, 2010) con-

tains a list of similarities between GFO and DOLCE

which helped with the alignment process. Some of

Table 1: Equivalence alignments for relational properties

between DOLCE-Lite and BFO; the alignments numbered

in bold font can also be mapped.

DOLCE-Lite BFORO

1. generic-location located in

2. generic-location-of location of

3. part has part

4. part-of part of

5. proper-part has proper part

6. proper-part-of proper part of

7. participant has participant

8. participant-in participates in

Table 2: Equivalence alignments for relational properties

between DOLCE-Lite and GFO; the alignments numbered

in bold font can also be mapped.

DOLCE-Lite GFO

1. generic-constituent has constituent part

2. generic-constituent-of constituent part of

3. generically-dependant-on depends on

4. generic-dependant necessary for

5. has-quale has value

6. quale-of value of

7. boundary has boundary

8. boundary-of boundary of

9. q-present-at exists at

10. temporary-participant-in agent in

11. temporary-participant has agent

12. generic-location occupies

13. generic-location-of occupied by

14. part abstract has part

15. part-of abstract part of

16. proper-part has proper part

17. proper-part-of proper part of

18. participant has participant

19. participant-in participates in

Table 3: Equivalence alignments for relational properties in

BFO and GFO; the alignments in bold are also mapped.

BFORO GFO

1. has part has part

2. part of part of

3. has proper-part has proper part

4. proper part of proper part of

5. has participant has participant

6. participant in participates

7. located in occupies

8. location of occupied by

9. has agent has agent

10. agent in agent in

the alignments could not be used, however, due to

changes in the two foundational ontologies. We were

able to use 42% of the alignments from the docu-

mentation. We discuss four equivalence alignments

from (Temal et al., 2010). We changed the align-

ment bfo:ProcessualEntities ↔ dolce:perdurant to

bfo:Occurrent ↔ dolce:perdurant, because by deﬁni-

AddressingIssuesinFoundationalOntologyMediation

tion occurrents and perdurants both represent entities

that have temporal parts and unfold in time. Temal

et al.’s alignment of bfo:Quality with dolce:physical-

quality is more precise than ours, because, as men-

tioned above, we chose to ignore the some philoso-

phies (the realist debate) with the hope of achieving

a higher number of alignments. That is, our mapping

has bfo:Quality ↔ dolce:quality, thereby ignoring the

fact that BFO does not consider abstract entities. We

agree with bfo:SpatialRegion ↔ dolce:space-region

and bfo:TemporalRegion ↔ dolce:temporal-region,

and use this equivalence, too. Seyed (Seyed, 2009)

examined only three relations—dependency, quality,

and constitution—and found that they are different in

DOLCE and BFO. The basic numbers of the align-

ments are included in Table 6.

Table 4: Equivalence alignments for classes between

DOLCE-Lite and GFO ontologies; the alignments num-

bered in bold font can also be mapped.

DOLCE-Lite GFO

1. particular Individual

2. endurant Presential

3. physical-endurant Discrete presential

4. physical-object Material object

5. amount-of-matter Amount of substrate

6. perdurant Occurrent

7. process Process

8. state State

9. abstract Abstract

10. set Set

11. quality Property

12. quale Property value

13. quality-space Value space

14. time-interval Chronoid

15. space-region Spatial Region

16. temporal-region Temporal Region

4.1.2 Automated Alignments

Table 7 lists the numbers of alignments found by the

selected tools. We describe some further data in the

remainder of this section.

H-Match generated many alignments, but most of

the output was not accurate. Many entity pairs that

were matched using H-Match were found to be incor-

rectly aligned; e.g., DOLCE-Lite:quale ↔ bfo:Role.

This resulted in us being able to use only 18% of

these alignments, with the rest being false positives.

PROMPT was generally unstable resulting in force

closure of the application. It generated suggestions

of which we were able to use 56%, with the rest being

false positives; e.g., bfo:Site ↔ gfo:Situoid.

While LogMap provided few alignments between

the foundational ontologies (less than ten in all

cases), most alignments were accurate. The one

Table 5: Common alignments between DOLCE-Lite, BFO

and GFO.

DOLCE-Lite BFORO GFO

Class

1. endurant Independent

Continuant

Presential

2. physical-object Object Material object

3. perdurant Occurrent Occurrent

4. process Process Process

5. quality Quality Property

6. space-region SpatialRegion Spatial region

7. temporal-region Temporal-

Region

Temporal region

Relational property

1. proper-part has proper part has proper part

2. proper-part-of proper part of proper part of

3. participant has participant has participant

4. participant-in participates in participates in

5. generic-location located in occupies

6. generic-

location-of

location of occupied by

Table 6: Comparison of manually performed alignment ac-

curacies of the GFO documentation (Herre, 2010), related

works, and ours, and aggregates for mappings.

Seyed Herre Temal

et al.

Ours

Class alignments

DOLCE-Lite ↔

BFO

- - 2/7 9/9

BFO ↔ GFO - - - 13/13

GFO ↔

DOLCE-Lite

- 13/31 - 16/16

Object property alignments

DOLCE-Lite ↔

BFO

0 - - 8/8

BFO ↔ GFO - - - 10/10

GFO ↔

DOLCE-Lite

- 0 - 19/19

Overall alignments

Total 0/0 13/31 2/7 75/75

Accuracy 0% 42% 29% 100%

Found 0% 37% 12% 100%

Overall mappings

Total 0/0 8/31 1/7 40/40

Accuracy 0% 26% 14% 100%

Found 0% 61% 9% 100%

false positive in LogMap was the alignment of

bfo:IndependentContinuant ↔ gfo:Independent.

YAM++ generated many alignments. However, while

most of the alignments for DOLCE ↔ BFO and BFO

↔ GFO were accurate, only about half were accurate

for GFO ↔ DOLCE. Overall we were able to use

almost 64% of its alignments. Like LogMap, YAM++

also incorrectly aligned bfo:IndependentContinuant

↔ gfo:Independent. Some of YAM++’s other false

positive alignments include dolce:generic-constituent

KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

Table 7: Comparison of alignment accuracies of the matching tools and aggregates for mappings.

H-Match PROMPT LogMap YAM++ Hot Match Hertuda Optima

Class alignments

DOLCE-Lite↔BFO 4/16 3/8 2/2 4/4 3/3 3/3 4/12

BFO↔GFO 5/31 7/12 7/8 6/7 7/7 7/7 8/14

GFO↔DOLCE-Lite 4/25 4/8 3/3 8/11 5/5 5/5 5/16

Object property alignments

DOLCE-Lite↔BFO 0 0 0 0 0 0 0/1

BFO↔GFO 0 0 4/4 0 0 0 1/3

GFO↔DOLCE-Lite 0 4/4 0 5/14 5/7 6/8 2/23

Overall alignments

Total 13/72 18/32 16/17 23/36 20/22 21/23 20/69

Accuracy 18% 56% 94% 64% 91% 91% 29%

Found 17% 24% 21% 31% 27% 28% 27%

Overall mappings

Total 10/72 11/32 16/17 15/36 11/22 12/23 13/69

Accuracy 14% 34% 94% 42% 50% 52% 19%

Found 25% 28% 40% 38% 28% 30% 33%

↔ gfo:has sequence constituent, dolce:quality-

space ↔ gfo:Space and dolce:temporary-proper-part

↔ gfo:has constituent part.

HotMatch generated a fair amount of alignments

between the ontologies. Overall, we were able to

use 91% of HotMatch’s alignments, with just 2 align-

ments out of all 22 being false positives. Hertuda’s

output was surprisingly similar to HotMatch’s output,

with just one more alignment than HotMatch. We

were able to use 91% of Hertuda’s alignments, with

just 2 alignments out of all 23 being false positives.

Common false positives in YAM++, Hertuda and Hot-

Match were the alignments between dolce:part ↔

gfo:has part and dolce:part-of ↔ gfo:part of, which

is discussed in Section 4.2.1. Optima generated many

alignments for each pair. However, there were many

false positives, consequently we were able to use 29%

of its alignments overall. This may be because Op-

tima is aimed at aligning large ontologies and the

foundational ontologies in question are of reasonable

size. Optima incorrectly aligned gfo:Continuous ↔

bfo:Continuant, dolce:Region ↔ bfo:SpatialRegion

and dolce:dependent-place ↔ bfo:Dependent.

4.2 Alignment Issues

We have encountered two types of issues in align-

ment: transitivity, where there was no ‘full circle’

alignment between some entities of the three ontolo-

gies, and approximate alignments, where there is no

clear relationship to describe the match.

4.2.1 Transitivity

Transitivity in entity alignments works as follows: if

the equivalence relation holds between entities from

the ﬁrst and second ontology and it holds between

entities from the second and third ontology; it nec-

essarily holds between entities from the ﬁrst and

third ontology. Applying transitivity to entity align-

ments assists in detecting errors. For instance, if

one were to align dolce:endurant ↔ gfo:Persistant,

and gfo:Persistant ↔ bfo:Continuant, then by tran-

sitivity this means that dolce:endurant is equivalent

to bfo:Continuant, which is incorrect. In most cases,

the foundational ontology alignments are transitive.

There were two types of exceptions, being the ab-

sence of an entity and what can be termed conse-

quences of conﬂicting philosophies.

Absence of an Entity. An alignment cannot be a

candidate for transitivity if there is an equivalence be-

tween only two out of the three ontologies. From the

three main ontology alignments, the following ones

were not transitive due to the absence of an entity:

• Absence of a DOLCE entity (7 cases): bfo:Entity

↔ gfo:Entity, bfo:DependentContinuant ↔

gfo:Dependent, bfo:ObjectBoundary↔gfo:Mate-

rial boundary, bfo:Function ↔ gfo:Function,

bfo:Role ↔ gfo:Role, bfo:has agent ↔

gfo:has agent, bfo:agent in ↔ gfo:agent in.

• Absence of a GFO entity (1 case): dolce:spatio-

temporal-region ↔ bfo:SpatioTemporalRegion

• Absence of a BFO entity (17 cases): gfo:Individual

↔ dolce:particular, gfo:Amount of substrate

↔ dolce:amount-of-matter, gfo:State ↔

dolce:state, gfo:Abstract ↔ dolce:abstract,

gfo:Set ↔ dolce:set, gfo:Property value ↔

dolce:quale, gfo:Value space ↔ dolce:quality-

space, gfo:Chronoid ↔ dolce:time interval,

gfo:has constituant part ↔ dolce:generic-

consitituant, gfo:constituant part of ↔

dolce:generic-constituant-of, gfo:necessary for

↔ dolce:generic-dependent, gfo:depends on ↔

AddressingIssuesinFoundationalOntologyMediation

dolce:generically-dependent-on, gfo:has value

↔ dolce:has-quale, gfo:value of ↔ dolce:quale-

of, gfo:has boundary ↔ dolce:boundary,

gfo:boundary of ↔ dolce:boundary-of,

gfo:exists at ↔ dolce:q-present-at

From this type of transitivity issue, we see that for the

three main ontology alignments, in most cases BFO

entities are absent. There are a few cases of absent

DOLCE entities and one case of an absent GFO entity.

Conﬂicting Philosophies. The philosophies of

foundational ontologies affect their entities to a cer-

tain extent. In some cases, two entities that are

aligned to each other may not be aligned to the same

entity of a third ontology.

• dolce:physical-endurant ↔ bfo:Material-

Entity, dolce:physical-endurant ↔ gfo:Discrete

presential and bfo:MaterialEntity ↔ gfo:Material

persistant. Let us align bfo:MaterialEntity ↔

dolce:physical-endurant, ignore their underlying

philosophies (i.e., that BFO is an ontology of

universals and DOLCE of particulars). However,

in GFO, there are two entities for representing

this type of entity, based on distinct philosophical

notions: gfo:Discrete presential, being subsumed

gfo:Individual, is suited for dolce:physical-

endurant while gfo:Material persistant, being

subsumed by gfo:Universal, is suited for

bfo:MaterialEntity.

• dolce:part ↔ bfo:has part, dolce:part ↔

gfo:has abstract part and bfo:has part ↔

gfo:has part (idem for their inverses). In

DOLCE, both the domain and range of part is

particular. In BFORO, there is no domain and

range for has part. In GFO, both the domain and

range of abstract has part is Item, while both the

domain and range for has part is Concrete. The

former relational property may be better suited

for DOLCE because it is a descriptive ontology

and contains abstract entities. The latter is better

suited for BFORO as it is a realist ontology,

representing the world as is, thereby containing

concrete entities only.

The ontology matching tools discussed in Sec-

tion 4.1.2 misaligned dolce:part ↔ gfo:has part and

their inverses. This is because object property incon-

sistencies are not fully recognised by reasoners (Keet,

2012), hence their conﬂicting domains and ranges did

not affect the satisﬁability of the ontology.

4.2.2 Approximate Alignments

There are a number of approximate alignments be-

tween foundational ontology entities. By this we

mean that they are not equivalent to each other or sub-

sumed by one another, but share some common char-

acteristics. By identifying these relations between

these entities, foundational ontology developers could

possibly relate them as sibling classes by grouping

them both under a common superclass. We mention

three of them.

• dolce:arbitrary-sum, bfo:ObjectAggregate and

gfo:Conﬁguration: All three of these entities de-

scribe a collection of something. dolce:arbitrary-

sum, however, has no unity criterion e.g., a pencil

and laundry basket are together a dolce:arbitrary-

sum, and it can contain both dolce:physical-

endurant and dolce:non-physical-endurant enti-

ties. dolce: physical-endurant is not restricted

just to instances of dolce:physical-object but can

possibly include dolce:feature and dolce:amount-

of-matter. bfo:ObjectAggregate, on the other

hand, has overall unity and can be considered

as a whole. It is restricted to bfo:Object only,

and in the case of BFO, all objects are physi-

cal. gfo:Conﬁguration is simply a collection of

gfo:Presential facts. gfo:Presentials are not re-

stricted to whole physical objects and can include

other gfo:Presential entities. For this reason, it

cannot equate to bfo:ObjectAggregate. Further-

more, it holds a restriction that it must contain

at least one material entity. dolce:arbitrary-sum

could contain physical, non-physical or both enti-

ties, with no restrictions.

• dolce:state and bfo:SpatioTemporalInstant:

DOLCE describes dolce:state by using an exam-

ple of a rock erosion describing state as a time in-

terval of the erosion is collapsed into a time point.

Similarly BFO deﬁnes bfo:SpatioTemporalInstant

as a “connected spatiotemporal region at a

speciﬁc moment”. The difference between the

two lies in the fact that dolce:state is homeomeric

while bfo:SpatioTemporalInstant is not.

• dolce:relevant-part and bfo:FiatObjectPart:

DOLCE describes dolce:relevant-part as a feature

that is a relevant part of their host; e.g., the edge

of a cube. BFO deﬁnes bfo:FiatObjectPart as a

material entity that is part of an object but not

demarcated by physical discontinuities; e.g.,

the lower portion of the leg. In this sense they

are both part objects that are physical entities.

However, it is unclear whether dolce:relevant-part

is demarcated by physical discontinuities or not

and whether BFO’s ﬁat object parts are ‘relevant’

somehow. This requires further investigation.

KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

5 MAPPING AND MERGING

Ontology Mapping uses the alignments from the

alignment process to create correspondences between

entities in the ontologies. The output from the align-

ment process is broader, while the output from the

mapping process is narrower as inconsistencies affect

the mapping process. Merging is performed by creat-

ing a new ontology of the source ontologies with their

mappings between each other. Ontology mapping and

merging was performed by relating classes and object

properties in Prot

e. Entities were mapped in the

order of their level in the hierarchy, from higher to

lower level, because foundational ontologies by deﬁ-

nition are general high-level ontologies. Therefore, in

mapping, preference must be ﬁrst given to high-level

entities to have agreement among general entities and

avoid inconsistencies at that level.

Alignments that cannot be mapped due to logical

inconsistencies result in unsuccessful mappings. The

inconsistencies were identiﬁed by using the following

method. For each candidate class mapping:

1. Assert the equivalence for the found alignment.

2. Run the automated reasoner.

3. Check if there are any unsatisﬁable classes.

4. If there are unsatisﬁable classes, use the reasoner

explanation feature to generate an explanation.

5. Analyse explanations.

6. Remove inconsistent mapping, if applicable.

For each candidate object property mapping, since

object property inconsistencies and ﬂaws are not

properly recognised by reasoners (Keet, 2012), we

identiﬁed inconsistencies by checking if an object

property pair’s domain and range restrictions are sat-

isﬁable by using the above method.

The numbers in bold face in Table 5 represent the

alignments that resulted in successful mappings be-

tween the common entities of the three main ontolo-

gies. From the 14 alignments in Table 5, six success-

ful mappings exist. Recall from the previous section

on alignment, there was a total of 85 distinct align-

ments between all foundational ontologies and related

modules, and 75 alignments between the main foun-

dational ontologies. Performing the method to iden-

tify inconsistencies in alignments resulted in 42 dis-

tinct logical inconsistencies of which 35 alone were

from the main ontologies. From all the distinct equiv-

alence alignments, only half were satisﬁable and re-

sulted in successful mappings. Comparing these map-

pings to the alignments found by the tools, LogMap

doubled its percentage found to 40% and performed

best compared to the six others evaluated (see Table 7,

bottom three rows).

To solve inconsistencies in the mapping attempts,

we analysed each alignment on the logical explana-

tion for the inconsistency and the description of the

entity provided by the foundational ontology develop-

ers, and checked whether it was possible to change the

alignment from equivalence to subsumption. How-

ever, there are still many unsolvable inconsistencies,

mainly due to hierarchical and structural differences

in the foundational ontologies. Due to space limita-

tions, we describe only a representative selection of

the logical inconsistencies and (logically satisﬁable)

possible solutions; the full list of inconsistencies is

available at http://www.thezﬁles.co.za/ROMULUS/.

Inconsistencies Due to Disjoint Classes. For this

type of inconsistency, the entities to be aligned are

disjoint to each other, either directly, through higher-

level equivalence relations or through their subclasses

or superclasses. If entities are disjoint, they cannot

overlap, hence cannot be equivalent.

• dolce:temporal-region - gfo:Temporal region -

bfo:TemporalRegion: The issue with incompat-

ible temporal regions between BFO, GFO, and

DOLCE is depicted in Fig. 1 and is a result of

the OWL DisjointClasses class axiom between

gfo:Concrete, gfo:Space Time and gfo:Abstract,

and between dolce:Abstract and dolce:Per-

durant, or, from the other viewpoint: because

BFO made TemporalRegion an Occurrent,

DOLCE made it Abstract, and GFO neither. This

does not seem to be resolvable.

• bfo:Role - gfo:Role: gfo:Processual role is a sub-

class of gfo:Role and gfo:Process. gfo:Process

is a subclass of gfo:Occurrent. gfo:Occurrent is

equivalent to bfo:Occurrent. bfo:Role is a sub-

class of bfo:Continuant. bfo:Continuant is dis-

joint to bfo:Occurrent. In this equivalence re-

lation, both gfo:Role and gfo:Occurrent are su-

perclasses of gfo:Processual role, and bfo:Role

is a subclass of bfo:Continuant; gfo:Occurrent

and bfo:Continuant are disjoint, hence the two

classes cannot be equivalent. Solution: Logi-

cally, bfo:Role cannot be equivalent to gfo:Role.

However, bfo:Role can be subsumed by gfo:Role.

Therefore the relation can be changed to gfo:Role

subsumes bfo:Role.

• gfo:necessary for - dolce:generic-dependent: If

we were to equate these object properties, we

would have to assume that their domains and

ranges are equivalent, which is not the case;

the situation is depicted in Fig. 2. Solution:

Logically, gfo:necessary for cannot be equiva-

lent to dolce:generic-dependent, because equat-

ing their domains and ranges causes inconsis-

tencies. However, dolce:generic-dependent’s

AddressingIssuesinFoundationalOntologyMediation

bfo:TemporalRegion

bfo:Occurent

gfo:Occurrent

gfo:Concrete gfo:Space_Time

gfo:Temporal_Region

≡

declaring equivalence results in inconsistency due to

disjointness among higher-level categories

gfo:Abstract dolce:Abstract

≡

dolce:temporal-region

dolce:Perdurant

≡

bfo:TemporalRegion

bfo:Occurent

Figure 1: Graphical depiction of why the aligned gfo:Temporal Region, bfo:TemporalRegion, and dolce:temporal-region cannot

be mapped in any way without causing an inconsistency; ≡: aligned entities, ×: disjoint entities.

gfo:Item

gfo:Category gfo:Individual

× ≡

declaring equivalence results in

inconsistency due to disjointness

and mappings among higher-

level categories

dolce:Particular

gfo:necessary_for

dolce:generic-dependent

Figure 2: Visualisation of the root cause of the

non-mappable gfo:necessary for and DOLCE-Lite:generic-

dependent; ×: disjointness, ≡: equivalence mapping.

domain and range, dolce:particular can logi-

cally be subsumed by gfo:necessary for’s do-

main and range, gfo:Item. Therefore the relation

can be changed to gfo:necessary for subsumes

bfo:generic-dependent.

• dolce:generic-location - bfo:located in: This

issue is due to disjointness among domain/range.

dolce:generic-location’s range is dolce:particular

and bfo:located in’s range is bfo:Continuant.

bfo:Continuant is disjoint to bfo:Occurrent

and bfo:Occurrent ≡ dolce:perdurant. In

DOLCE, perdurant v ∃has-Quality.temporal-

location-q and the domain of dolce:has-

Quality is dolce:particular (the superclass of

dolce:perdurant). Thus, bfo:Continuant is dis-

joint to ∃has-Quality.temporal-location-q, and

cannot be equivalent, causing the range re-

strictions to be unsatisﬁable in the alignment.

Therefore dolce:generic-location cannot map to

bfo:located in.

Other unresolvable cases are, among others, dolce:set

- gfo:Set and dolce:quality-space - gfo:Value space.

Inconsistencies Due to Complement Classes. For

this type of inconsistency, the entities to be aligned

were found to be complements of each other, either

directly, through higher-level equivalences or through

subsumption. We describe here one such case.

• bfo:MaterialEntity - gfo:Material

persistant,

which is visualised in Fig. 3. The cru-

cial aspect in GFO is the class axiom

Universal v ∃instantiated by.Item, and the com-

plement for individuals. Concerning mappings,

bfo:IndependentContinuant ≡ gfo:Presential.

gfo:Material_persistant

gfo:Universal gfo:Item

≡

declaring equivalence results in

inconsistency due to the complement

bfo:IC

instantiated_by some

gfo:Individual

NOT instantiated_by some

gfo:Presential

bfo:MaterialEntity

Figure 3: Visualisation of the root cause of the non-

mappable bfo:MaterialEntity and gfo:Material persistant; ≡:

equivalence mapping, IC = IndependentContinuant.

However, GFO has Presential v Individual

and Individual v ¬∃instantiated by.Item.

Thus, gfo:Material persistant is a subclass

of gfo:instantiated by some gfo:Item while

bfo:MaterialEntity is a subclass of the com-

plement of that class, hence bfo:MaterialEntity

cannot be equivalent to gfo:Material persistant.

Solution: The alignment can be changed into

bfo:MaterialEntity - gfo:Discrete presential,

which avoids the complement issue but it is

not free of argument (recall the “conﬂicting

philosophies” item in Section 4.2.1).

6 DISCUSSION

Given the size of the ontologies and our high toler-

ance by ignoring underlying philosophies, the amount

of alignments, and, even more so, the amount of map-

pings is less than one may have expected; or: once

investigated in detail, the foundational ontologies are,

at present, not particularly interchangeable even at the

logical level. Only six pairwise mappings exist, i.e.,

they being, essentially, equivalent throughout all three

examined foundational ontologies.

Concerning feasibility to carry out automated

alignments, in most cases, the tools evaluated with

the OAEI performed better than the others, with the

exception of Optima. LogMap had the highest ac-

curacy, because it also considers the logic-based se-

mantics of the ontologies and uses automated reason-

ing services throughout the process, therewith elim-

inating those false positives that would have led to

a logical inconsistency. However, LogMap gener-

KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

ated very few alignments compared to other accu-

rate tools (YAM++, Hertuda and HotMatch), indicat-

ing that the additional heuristics implemented are too

strict at least for foundational ontology alignment.

Most false positive alignments generated by

the tools, such as bfo:IndependentContinuant ↔

gfo:Independent, indicate that the algorithms imple-

ment syntactic matching, which, based on the results

we obtained, is not sufﬁcient or suitable for foun-

dational ontology matching because many entities

have a common syntax e.g., dolce:quality-space ↔

gfo:Space both have the string ‘space’ in common but

are entirely different entities; Table 8 includes a selec-

tion of such false positives that are caused by syntactic

matching in the tools when aligning the three founda-

tional ontologies. The tools failed to recognise simple

alignments such as dolce:perdurant ↔ gfo:Occurrent,

bfo:Quality ↔ gfo:Property. In this sense, semantic

matching is not considered, or if it is, it fails to recog-

nise synonyms of the philosophical scope on which

foundational ontologies are built upon. Structural

matching is not an effective method either, due to the

fact that the hierarchies and structures of the founda-

tional ontologies differ greatly which causes the root

distances of mappable entities to differ. For aligning

foundational ontologies, it will be useful if existing

semantic matchers would include something alike a

‘philosophy WordNet’ that specialises in philosophi-

cal terms, synonyms, and deﬁnitions used in founda-

tional ontologies.

Table 8: False positives caused by syntactic matching gen-

erated by the alignment tools; the terms in italics represent

the strings that are common between aligned entities.

DOLCE-Lite BFO

physical-region ConnectedSpatio

TemporalRegion

non-physical-object Object

region SpatioTemporalRegion

BFO GFO

IndependentContinuant Independent

Site Situoid

Continuant Continuous

GFO DOLCE-Lite

has sequence

constituent

generic-constituent

has-part part

Space quality-space

The results of the tool analysis is a good indica-

tion of which tools to experiment with for founda-

tional ontology alignment in general. However, they

found less than a third of the actual alignments at this

stage, and therefore it is still vital to perform manual

alignment for foundational ontologies. The tools also

did not generate subsumption relations for any of the

alignments, but this could perhaps be an extension to

the basic idea of LogMap by means of another call

to the reasoner. One could investigate whether Op-

tima is useful to identify accurate alignments among

the larger foundational ontologies SUMO (Niles and

Pease, 2001) and YAMATO (Mizoguchi, 2010).

On a positive note, the systematised list of issues

now can be taken up by ontologists. While some of

the inconsistencies found are quite elaborate, others

should be easier to resolve both ontologically (philo-

sophically) and where in the ontology the entity is

positioned; e.g., the notion of a mathematical Set is

fairly well investigated already, and likewise the dif-

ferent theories of parthood. As such, the results pre-

sented here provide a solid foundation for ample on-

tological investigations. From an engineering view-

point and in case of urgent need for interoperability,

one could take a quite different strategy: OWL 2 EL

does not have negation, and therefore it should be pos-

sible to assert more mappings between the OWL 2 EL

modules of the foundational ontologies. Whether that

is the best strategy is a different matter, and it does not

take away the substantial list for which there was no

transitivity due to ‘missing’ entities. In any case, we

now know that some mappings are possible, hence,

also some foundational ontology interoperability.

7 CONCLUSIONS

The foundational ontologies DOLCE, BFO, and GFO

were pairwise aligned and mapped. They were

aligned manually, which served as the ‘gold stan-

dard’, and with the aid of seven alignment tools. The

accuracy and percentage of alignment found were

compared, where LogMap had the highest accuracy

with 94% and HotMatch and Hertuda as close sec-

ond, and YAM++ found the most correct alignments

(31% of the total manual alignments among the three

main ontologies (75)). The evaluation of the tools in-

dicated that the algorithms currently implemented by

the tools are not well-suited for foundational ontology

mediation. Semantic matching in the tools need to be

improved to include philosophical synonyms which

are used in foundational ontologies. Declaring the

correspondences in all ontology ﬁles based on its 85

alignments resulted in only 43 mappings, with the re-

maining 42 causing logical inconsistencies. The in-

consistencies are due primarily to differences in their

respective hierarchical structure with conﬂicting ax-

ioms, such as complement and disjointness, and in-

compatible domain and range restriction. On closer

inspection, some inconsistencies may be resolved us-

ing subsumption or making them sibling classes.

AddressingIssuesinFoundationalOntologyMediation

Future research includes mapping other founda-

tional ontologies, adding subsumption mappings, and

evaluating the current alignments with the founda-

tional ontology developers. We also aim to imple-

ment a facility for community input on the alignments

and mappings, which could to be facilitated via the

foundational ontology library that is available online

at http://www.thezﬁles.co.za/ROMULUS/.

REFERENCES

Castano, S., Ferrara, A., and Montanelli, S. (2003). H-

MATCH: an algorithm for dynamically matching on-

tologies in peer-based systems. In The ﬁrst Inter-

national Workshop on Semantic Web and Databases

(SWDB’03), pages 231–250, Humboldt-Universitt,

Berlin, Germany.

Dang, T. T., Gabriel, A., Hertling, S., Roskosch, P.,

Wlotzka, M., Zilke, J. R., Janssen, F., and Paulheim,

H. (2012). HotMatch results for OEAI 2012. In Sev-

enth International Workshop on Ontology Matching

(OM’12), volume 946 of CEUR Workshop Proceed-

ings. CEUR-WS.org.

de Bruijn, J., Ehrig, M., Feier, C., Mart

ıns-Recuerda, F.,

Scharffe, F., and Weiten, M. (2006). Ontology medi-

ation, merging, and aligning. In Semantic Web Tech-

nologies, pages 1–20. Wiley Online Library.

Euzenat, J. and Shvaiko, P. (2007). Ontology Matching.

Springer.

Herre, H. (2010). General Formal Ontology (GFO): A foun-

dational ontology for conceptual modelling. In The-

ory and Applications of Ontology: Computer Appli-

cations, chapter 14, pages 297–345. Springer.

Hertling, S. (2012). Hertuda results for OEAI 2012. In Sev-

enth International Workshop on Ontology Matching

(OM’12), volume 946 of CEUR Workshop Proceed-

ings. CEUR-WS.org.

Jim

enez-Ruiz, E. and Cuenca Grau, B. (2011). LogMap:

Logic-based and scalable ontology matching. In 10th

International Semantic Web Conference (ISWC’11),

volume 7031 of Lecture Notes in Computer Science,

pages 273–288. Springer.

Keet, C. M. (2012). Detecting and revising ﬂaws in OWL

object property expressions. In ten Teije et al., A., ed-

itor, 18th International Conference on Knowledge En-

gineering and Knowledge Management (EKAW’12),

volume 7603 of Lecture Notes in Artiﬁcial Intelli-

gence, pages 252–266. Springer. Oct 8-12, Galway,

Ireland.

Khan, Z. and Keet, C. M. (2012). ONSET: Automated foun-

dational ontology selection and explanation. In ten

Teije et al., A., editor, 18th International Conference

on Knowledge Engineering and Knowledge Manage-

ment (EKAW’12), volume 7603 of Lecture Notes in

Artiﬁcial Intelligence, pages 237–251. Springer. Oct

8-12, Galway, Ireland.

Khan, Z. and Keet, C. M. (2013). Toward semantic in-

teroperability with aligned foundational ontologies in

ROMULUS. In Seventh International Conference on

Knowledge Capture (K-CAP’13). ACM proceedings.

23-26 June, Banff, Canada. (poster/demo, in press).

Kolli, R. and Doshi, P. (2008). OPTIMA: tool for ontology

alignment with application to semantic reconciliation

of sensor metadata for publication in SensorMap. In

IEEE International Conference on Semantic Comput-

ing (ICSC’08), pages 484–485. IEEE. 4-7 Aug. 2008,

Santa Clara, CA, USA.

Masolo, C., Borgo, S., Gangemi, A., Guarino, N., and

Oltramari, A. (2003). Ontology library. Won-

derWeb Deliverable D18 (ver. 1.0, 31-12-2003).

http://wonderweb.semanticweb.org.

Mizoguchi, R. (2010). YAMATO: yet another more ad-

vanced top-level ontology. In Proceedings of the Sixth

Australasian Ontology Workshop, Conferences in Re-

search and Practice in Information, pages 1–16. Syd-

ney : ACS.

Ngo, D. and Bellahsene, Z. (2012). YAM++ : A multi-

strategy based approach for ontology matching task.

In ten Teije et al., A., editor, 18th International

Conference on Knowledge Engineering and Knowl-

edge Management (EKAW’12), volume 7603 of Lec-

ture Notes in Artiﬁcial Intelligence, pages 421–425.

Springer. Oct 8-12, Galway, Ireland. Demo.

Niles, I. and Pease, A. (2001). Towards a standard upper

ontology. In Second International Conference on For-

mal Ontology in Information Systems (FOIS’01). IOS

Press. Ogunquit, Maine, October 17-19, 2001.

Noy, N. F. and Musen, M. A. (2000). PROMPT: Al-

gorithm and tool for automated ontology merging

and alignment. In Seventeenth National Conference

on Artiﬁcial Intelligence and Twelfth Conference on

on Innovative Applications of Artiﬁcial Intelligence

(AAAI/IAAI), pages 450–455. AAAI Press.

Seyed, A. P. (2009). BFO/DOLCE primitive relation com-

parison. In The 12th Annual Bio-Ontologies Meeting

Colocated with Intelligent Systems for Molecular Bi-

ology (ISMB’09). Stockholm, Sweden, 28 June 2009.

Smith, B., Ceusters, W., Klagges, B., Kohler, J., Kumar, A.,

Lomax, J., Mungall, C., Neuhaus, F., Rector, A., and

Rosse, C. (2005). Relations in biomedical ontologies.

Genome Biology, 6(5):46.

Temal, L., Rosier, A., Dameron, O., and Burgun, A. (2010).

Mapping BFO and DOLCE. Studies in Health Tech-

nology and Informatics, 160(Pt 2):1065–9.

KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment