Conference PaperPDF Available

Evolution of the Metadata in the Ontology-based Knowledge Management Systems.

Authors:

Abstract and Figures

An ontology-based knowledge management system uses an ontology to represent explicit specification of a business domain and to serve as a backbone for providing and searching for knowledge sources. But, dynamically changing business environment implies changes in the conceptualisation of a business domain that are reflected on the underlying domain ontologies. Consequently, these changes have effects on the performances and validity of the KM system. In this paper we present an approach for enabling consistency of the description of knowledge sources in an ontology-based KM system in the case of changes in the domain ontology. This approach is based on our research in the ontology evolution and ontology-based annotation of documents. The proposed method is implemented in our semantic annotation framework so that efficient acquiring and maintaining of ontology-based metadata is supported.
Content may be subject to copyright.
Evolution of the Metadata in the Ontology-based
Knowledge Management Systems
Ljiljana Stojanovic1, Nenad Stojanovic2, Siegfried Handschuh2
1FZI Research Center for Information Technologies
at the University of Karlsruhe, Germany
2Institute AIFB - University of Karlsruhe, Germany
Abstract: An ontology-based knowledge management system uses an ontology
to represent explicit specification of a business domain and to serve as a
backbone for providing and searching for knowledge sources. But, dynamically
changing business environment implies changes in the conceptualisation of a
business domain that are reflected on the underlying domain ontologies.
Consequently, these changes have effects on the performances and validity of
the KM system.
In this paper we present an approach for enabling consistency of the description
of knowledge sources in an ontology-based KM system in the case of changes
in the domain ontology. This approach is based on our research in the ontology
evolution and ontology-based annotation of documents. The proposed method
is implemented in our semantic annotation framework so that efficient
acquiring and maintaining of ontology-based metadata is supported.
1. Introduction
In the dynamically changing world of business, the competitiveness of companies
depends heavily on the possibility to find, for a given problem, the right knowledge in
the right moment. This view presumes existence of knowledge sources and focuses on
the acquiring, using and validating knowledge sources – so called supply-side of the
knowledge management system [Mc01]. Practically, such KM approach is based on
the knowledge integration process [Fi01], in which heterogeneous forms of
knowledge sources (text, audio, video) should be integrated through unified searching
interface in order to find right solution for a given problem. Expansion in the using
Internet technologies for corporate IS implies using web portals as unique interface
for providing and accessing content of various knowledge sources. The prerequisite
for this integration is the unified description of the content of the knowledge sources –
unified format and used vocabulary. As a promising integration structure appears
ontologies that provide real-world and formal semantic of a domain theory. There are
several approaches for using ontologies in KM [Fe00]. Ontology-based KM systems
use ontologies as a backbone for providing and accessing knowledge sources. An
ontology offers a terminology for the knowledge indexing and searching process. The
main advantage comparing to key-word based indexing/searching is that an ontology
is a formalised, common and shared description of a domain. Therefore, it provides a
set of assumptions about intending meaning of used terms, e.g. when one searches for
a knowledge source that is about the animal “jaguar” then it is avoided to retrieve a
source that describes the “jaguar” car.
An ontology in a KM system is related to the business strategy and also indirectly to
the business environment. Consequently in a fast changing environment it is obvious
that an ontology as a domain backbone is also a matter of change. The changes have
to be propagated to all descriptions, e.g. annotations of the knowledge sources in
order to enable consistency. Although this change propagation problem has great
impact on knowledge searching process, this problem is not well addressed in the KM
literature [Ha00].
In this paper we present an approach that enable consistency in the annotations of
knowledge sources in the case of changes in the domain ontology. The approach is
based on our research in the area of ontology evolution and ontology-based
annotation of documents [Ha01]. The proposed method is included in our semantic
annotation framework CREAM so that efficient acquiring and maintaining of
ontology-based metadata is supported.
The benefits of the proposed approach are manifold:
- In the case of changes in the domain ontology, annotations of knowledge sources
can be automatically updated;
- An ontology-evolution model enables the categorisation of required/derived
changes so that incorrectnesses which lead to the more critical decreasing of the
system’s performances can be managed firstly;
- A special ontology for maintenance of the annotation is introduced -- herein after
called maintenance ontology. It offers new search possibilities for knowledge
sources, not only according to the content, but also according to author, date,
format, relevance and their combinations.
From the knowledge management system point of view the proposed approach will
enable us to develop a robust knowledge management solution that copes with the
high-changeable business conditions.
Paper is organised as follows: Section 2 describes the typical problems of the
knowledge-management systems in the dynamically changing world of business.
Section 3 explores the problem of changes in the ontology and analyses the effect of
the change on the ontology itself and on the underlying objects. In section 4 we
describe a method to analyse and propagation changes made in the ontology. Further
we present an integration tool for implementing this method in a KM scenario. Before
we conclude, we give a survey of related work in the categories knowledge
management, ontology evolution and annotation environments.
2. Maintenance problem in an ontology-based Knowledge
Management system
The frequency and variety of doing a business implies the production of tons of
information in various representation formats (text, audio, video) and various levels of
structures (structured, semi-, un-structured). Information are spread all overall a
company: in business documents, technical documentations, manuals, legacy
databases, and e-mails. Most of these items could be treated as very valuable
knowledge sources for a particular problem. One of the most important tasks of a
knowledge management system is to find effectively the appropriate content in all of
these heterogeneous sources. All the systems of the first generation of KM [Mc01] are
supply-side, which means that they are focused mainly on searching for relevant
knowledge sources (knowledge integration process) and less (or not at all) on the
knowledge production process [Fi01].
Using Internet infrastructure enhances in many ways “searching for knowledge”
practices of an KM system [OL98], while the Web provides a possibility to integrate
all of these sources on the presentation level: all of them could be presented to the
user through a single interface – the Web browser. Therefore, a human expert could
use the same Web interface (for example, searching engines like AltaVista.) to find
relevant information stored in the text, pictures, video files. But, real experience
teaches us that such an expert should not be only a domain expert (for example bio-
chemist) but a searching-expert (for particular searching engine) as well.
The mentioned searching-problem lies in the structure of the current Web – it is
designed only for human consumption - machines so far mainly help in better
presenting of information, but with limited possibilities to process the content of the
presented information. Therefore, the real integration of the knowledge sources has to
be done by the formal introduction of an intermediate level (between syntax and
presentation levels) that will help software agents to understand the content of the
knowledge sources.
This machine understanding assumes [Fe00]:
- a formal understanding that allows the processing of the semantic by a computer;
- a real-world understanding that allows relating semantic of information to the
common-shared meaning of humans.
It can be realised by annotating each knowledge source with a formal description of
the content.
As a promising structure for realising such a machine understanding appears
ontologies, an explicit specification of the conceptualisation of the domain of interest
[Gr93]. Ontologies typically consist of definitions of concepts relevant for the
domain, their relations, and axioms about these concepts and relationships. They
provide a suitable format and a common-shared terminology for the description of the
content of knowledge sources. In other words, each knowledge source should be
semantic annotated, i.e. enriched with a metadata description [Ha01].
The semantic annotation resolves one of the common problems in the underlying
ontology-based KM systems: the prediction game between indexers and users. An
indexer attempts to predict which concepts a user will employ when searching for a
particular knowledge source. In formulating a query, the user attempts to predict
which index concepts are attached to the knowledge source he or she seeks. By using
a given domain ontology one can annotate content of provided knowledge source in
such a way that a knowledge-seeker can find that knowledge source easily,
independently of its representation format– which is the vision of an ontology-based
supply-side KM system [St01].
The suggested changes in the machine-understandable description of the content of
the knowledge sources in the Web require changes in the basic WWW infrastructure,
which leads to the second generation of the Web – so called Semantic Web [BL00].
The basic infrastructure for the Semantic Web is on the way and the presented KM
scenario could be one of the “killer applications” for the Semantic Web.
However, there are several crucial problems, which should be resolved in order to
realise this “KM dream”. We mention only a few: how to define an ontology, how to
support semantic annotation of the heterogeneous (audio, video) knowledge sources,
how to define a query language on the metadata level. But from the point of view of a
KM system the most important question arises: is the metadata assigned to a
knowledge source valid, i.e. up-to-date?
As mentioned in the introduction, each KM system should reflect indirectly, implied
by changes in the ontology, all the changes made in the business environment.
Particularly, each change in the business conceptualisation (changes in the strategy of
the company, in the market planning, in the customer segmentation) requires changes
in the domain modelling and each change in the domain modelling should be reflected
in the changes of the metadata description of the content of the knowledge sources.
That does not mean that every change in business environment implies a change in
the validity of knowledge sources - knowledge sources are valid but the view on the
content of these knowledge sources is changed. From the annotation point of view the
terminology used for the description of the content of the knowledge sources should
be changed according to the business changes.
A knowledge management system could decrease its efficiency drastically in the case
that some of the knowledge sources are annotated with an “old” ontology and that a
“revised” ontology is used for searching. For a given query it would not only miss
some relevant knowledge sources, but also deliver WRONG answers, for example in
the case of the query: “Give me all knowledge sources that describe bonuses of our
customers” given in figure 2.1.
Figure 2.1. Incorrect statement – part of business ontology: Concept “customer” is divided
in two subconcepts: “Privileged” and “others”. “Privileged” customers are divided into
“software” (only software is sold to them) and “hardware” (only hardware is sold to them). In
the “old” ontology (left side) concepts “software” and “hardware” inherit a property “bonus”
from their parent. In the “new” ontology (right side) the concept “hardware” changed its parent.
However, the new parent does not contain the property “bonus”. As a consequence, the
knowledge source “Y”, that is about bonus for customer who buy hardware, is incorrect while
in the “new” ontology (new business policy) this type of customer has no bonus privilege. The
meaning of used representation formalism is detailed in the section 3
In the next sections we describe a method (section 3, 4) to analyse and propagation
changes made in the ontology. Further we present an integration tool (section 4) for
implementing this method in a KM scenario.
3. Ontology evolution
One critical point in applying ontologies to real-world problems is that domains are
changing fast (new concepts evolve, concepts change their meaning, new business
rules are defined, etc.) and user needs are changing, too. Thus, the corresponding
ontologies have to evolve as well. Ontology evolution is the timely adaptation of the
ontology to the changed business requirements, to the trends in the ontological
instances and to the way of using of the ontology-based applications, as well as the
consistent management/propagation of these changes because a modification in one
part of the ontology may generate subtle inconsistencies in other parts of the same
ontology, in the ontology-based instances, depending ontologies and applications.
This variety of causes and consequences of the ontology changes makes ontology
evolution a very complex process (figure 3.1) that is described in the following.
Figure 3.1. Phases in the ontology evolution process
Elementary changes in the ontology are shown in the table 3.1, including the addition
and the deletion of all ontological entities. Modification (update) of any ontological
entity is realized using the deletion of the old entity and the addition of the new entity.
The single exception is the modification in the concept hierarchy because of the
relation’s inheritance. Change in the name of the entity is not considered, because
every entity has unique identifier that is independent of the entity name.
Table 3.1. Elementary changes in the ontology
Elementary change
Add AddConcept, AddRelation, AddIsA, AddAxiom, AddDomain, AddRange
Delete DeleteConcept, DeleteRelation,DeleteIsA,DeleteAxiom,DeleteDomain, DeleteRange
Modify Modify_IsA
3.1 Semantics of change
An ontology has to be consistent according to its structure (concepts, inheritance
graph, relations, axioms). This is “semantics of change” phase (cf. figure 3.1 (A)) that
refers to the effect of the change on the ontology itself. In order to retain consistency
of the ontology, set of required changes is expanded with the additional (derived)
changes in ontology. For example, the deletion of relation domain can provoke the
deletion of the relation as well in the case that there are no other concepts defined as
domain of this relation.
The additional changes in the ontology are derived automatically. The approach is
based on the sound and complete set of axioms (provided with an inference
mechanism) that formalises the dynamic of the ontology evolution. The compliance
of the available ontology changes with the axioms automatically ensures ontology
consistency, without need for explicit checking as incorrect ontology version cannot
actually be generated [Fr00]. While the focus of the paper is on the knowledge
management, we will omit here the description of our approach used for “semantics
of change” and concentrate on the “change propagation” problem, which has great
impact on the knowledge searching process.
Inputs of this phase (A) are required changes (1) and source ontology (2a) and outputs
are list of required and derived changes (3) and modified source ontology (2b).
3.2 Change propagation
Potentially, an ontology change might corrupt the instances, dependent ontologies as
well as application programs running against the ontology and/or the knowledge base.
The task of the change propagation phase is automatically bringing all dependent
elements to a consistent state after an ontology update has been performed. Block (B)
in the figure 3.1 depicts this phase. Output is the list of changes (7), which have to be
done. In the rest of the section we will analyse the effect of the change propagation on
the corresponding inputs.
Effect of changes on the dependent ontologies
An ontology update might corrupt ontologies that depend on the modified ontology.
They are built from the modified ontology or they import it. This problem could be
solved by recursive applying ontology change procedure on these ontologies in order
to preserve their conceptual, structural and behavioural consistency [Fr00].
Effect of changes on the ontological instances
When the ontology is modified, the instances need to be changed in such a way that
the ontology and instances remain consistent with each other. Basically, if the
ontology is modified instances must be transformed to confirm to the modified
ontology. It means that continuous adaptation of the annotated information to the new
semantic terminology and relationships is necessarily.
Effect of changes on the applications
Changes in the ontology might invalidate applications that are already running on top
of the ontology and the knowledge bases, especially if they rely on certain schema
characteristics, which are lost after the ontology update. In the ideal case, the
conceptual knowledge that is necessary for an application should be merely specified
in the ontology. However, practice applications also use an internal model that may
become incompatible with the ontology [KF01]. Moreover, although the application’s
programs are written to be as generic as possible, there are a certain number of “hard-
coded” elements that should be treated special in some way. In most of the web
applications, where some queries are “hard-coded” into the service that is invoked as
a response on the specific action, the query rewriting process is needed [FL96].
4. Evolution of the metadata
This section introduces the backbone of our approach - evolution ontology that
supports, alleviates and automates the evolution process. Thereupon, we present our
method for solving the change propagation problem and our annotation framework,
which integrates the ontology evolution process.
4.1 Evolution ontology
Since ontology evolution requires additional meta-level reasoning capabilities that
allow inspecting changes and their logical dependencies, we define a special
ontology, so-called evolution ontology. We distinguish between domain ontology that
is changed and the evolution ontology that enables better management of these
changes.
Ontological changes are represented using the top level concept "Change", its
subconcepts (AddConcept, AddRelation, etc. ) and its relations [Ol99]. For every
change, it is also useful to know who is author of the change and when it is happened
(time). The cause of the change is used to represent the source of the change (business
requirements or the learning process) and the relevance of the change describes
whether and how it can fulfil the requirements. Also, ontology evolution is
managerial process and it needs some properties to support decision-making like cost,
priority, etc. Order of the changes is also very important while it enables recovery of
implemented changes. Moreover, change propagation cannot be done after every
change in the ontology (it requires too much time) even thought the change causes
instance inconsistency. Consequently, only the order of the changes can guaranty that
the instances “picture” the real status of the ontology structure. To solve semantics of
change problem, the evolution ontology contains axioms that derive additional
changes. Similarly to the ordering of the change: this type of the dependency between
changes is represented as a relation parentChange.
The second part of the evolution ontology represents semantic information about the
domain ontology explicitly (relations prototypical, primary_key, etc.), because the
conceptual structure of the evolution ontology aims to provide enough mechanisms to
deal with problems of syntax as well as semantic inconsistencies that arise when the
domain ontology is changed [TB01]. The third part of the evolution ontology aims to
support data-driven self-improvement of the domain ontology. For example, the fact
that these are no instances of some concept is a sign that this concept should be
deleted. We enforce formal discovering of changes by representing these heuristics as
axioms in the evolution ontology.
The benefits of using the evolution ontology are manifold: First, changes are formally
represented. Second, a history of changes is stored. Third, based on the formal
representation and the history of changes the change-propagation problem may be
approached. Using the same representation model for the ontology and analysis of
changes simplifies storage and allows reuse of system components like searching.
4.2 Evolution of the metadata
In this section we present our method for the change propagation problem based on
consistency analysis of already existing metadata and the performed change in the
domain ontology. It is divided into three steps described in the following.
Metadata capturing
When an ontology is modified, instances need to be changed in such a way that the
ontology and instances remain consistent with each other. If the instances are on the
Web, they are collected in the knowledge base using tools like focused crawler1
(process “capture” in the figure 4.1). In order to speed up the whole change
propagation process, only the instances that depend on the change are gathered. This
dependency information is obtained from the instance of the evolution ontology that
represents the performed change. Moreover, the output of this step is one list that
makes references between located instances and Web documents.
The main problem is how to find an application that uses the ontology that is changed.
An application can be semi-automatic maintained only if exists metadata describing
which ontology and/or ontological entities that application uses. Thus, annotation of
applications is necessary.
Metadata analysis
In the second step, automatic translation of the instances is performed according to
the changes in the ontology [SSV02]. In order to avoid overhead of the system, which
may heavily increase if the changes are performed every time the ontology has to be
modified, the categorisation of the changes is embedded in the evolution ontology.
We distinguish between:
- ontology-extending changes that do never have an impact on the existing
instances (e.g. creating a new relation);
- changes that provoke syntax inconsistencies in the ontological instances (e.g.
deleting a concept that already has instances);
- changes that provoke semantic inconsistencies in the ontological instances (e.g.
creating a new sibling concept does not lead to the invalidity in the set of
instances but an analysis of the meaning of the instances is needed).
The axiomatic part of the evolution ontology enables the verification of the formal
characteristic of the instances. The analysis of the semantic consistency is based on
the meta information (e.g. primary_key) defined in the evolution ontology.
This step provides an output in the form of list of modified instances with the
reference to the corresponding resource (knowledge source). Only this step is
performed in the case that instances are already gathered in the knowledge base.
Generation of a proposal for modifications
In the last step “out of date” instances on the Web are replaced with the corresponding
“up-to-date” instances. As depicted in the figure 4.1, some modifications of the
instances can be done automatically (process “update”), but for the instances that are
“write-protected” the notification has to be sent (process “notification) to the author
of the annotation in order to inform her/him about the changes and to suggest how to
correct the instance. Information about author is saved in the property “Author” in the
evolution ontology.
Using the method for metadata evolution does not solve all problems. However, we
provide guidelines, which suggest which resources’ metadata have to be checked, and
eventually changes to run again the changed ontology.
1 Kaon.semanticweb.org
4.3 Framework
In order to support the proposed approach for ontology evolution based on the
maintenance of the instances we have adapted our CREAM framework [Ha01]
presented in the figure 4.1. The Evolution Ontology, Evolution Component and
related links are the new elements and they are described in the previous section.
Figure 4.1. Architecture of CREAM
Document Editor/Viewer: The document editor/viewer visualizes the document
content and the annotations.
SOEP2 - Ontology and Fact Editor: The user can browse and edit the ontology and
retrieve for one concept all instances or for one instance all properties.
Crawler: The creation of relational metadata must take place within the Semantic
Web. During metadata creation subjects must be aware of which entities exist already
in their part of the Semantic Web. This is only possible if a crawler makes relevant
entities immediately available.
Annotation Inference Server: Relational metadata, proper reference and avoidance
of redundant annotation require querying for instances, i.e. querying whether and
which instances exist. For this purpose as well as for checking of consistency, we
provide an annotation inference server in our framework. The annotation inference
server reasons on crawled and newly created instances and on the ontology. It also
serves the ontological guidance and fact browser, because it allows querying for
existing concepts, instances properties.
Document Management: In order to avoid redundancy of metadata creation efforts,
it is not sufficient to ask whether instances exist at the annotation inference server.
When a metadata creator decides to capture knowledge from a Web page, he does not
want to query for all single instances that he considers relevant on this page, but he
2 kaon.semanticweb.org
wants information, whether and how this Web page has been annotated before.
Considering the dynamics of HTML pages on the web, it is desirable to store
annotated web pages together with their annotations. When the web page changes, the
old annotations may still be valid or they may become invalid. The metadata creator
must decide based on the old annotations and based on the changes of the web page.
5. Related Work
Knowledge management and annotation/ontology evolution
As known to authors the problem of maintaining description (annotations) of
knowledge sources in an ontology-based KM system in the case of changes in the
domain ontology is not treated in the literature and therefore we here present an
analyse of the annotation systems for the knowledge management purposes. The last
presented system gives the best view on the maintenance problem in the knowledge
management community. Annotate [Gi99] is a system that use information retrieval
methods to support KM in an organisation. It enables document annotations on the
web and captures global usage history. Annotate is not ontology-based and therefore
does not treat the problem of managing validity such knowledge item descriptions. In
[DPP00] paper author presents several issues with the design and implementation of
organisation memories in distributed companies. They have designed a tool, based on
the domain model in the form of ontology, capable to capture the content of the
documents and the context, in which they were created. A sophisticated retrieval
engine can retrieve the annotated documents based on their context. The presented
system seems very similar to ours; it has very suitable user interface which support
process of creating document annotations, it is integrated in the general ontology
engineering environment, but it is not adapted to new web infrastructure (Semantic
Web) and does not consider ontology evolution problem.
A very interesting, field research study of managing changes in a knowledge
management system is given in [Ha00]. The authors consider two types of changes:
(i) functional changes that are about new KM-systems in the organization, new
versions of a KM-system and new features in one KM-system and (ii) structural
changes that deal with new business models, new subsidiaries and new competencies
in the organisation. The results of the study show that managing the evolution of KM-
systems on an ad hoc basis can lead to unnecessary complexity and KM-systems
failures and that KM research has paid little attention to the evolution of KM-systems.
Ontology evolution
There are very few approaches investigating the problems of changing in the
ontologies. The most similar approach to our approach is described in the paper
[KF01]. As the authors also mentioned the most important flaw is the lack of a
detailed analysis of the effect of specific changes on the interpretation of data.
The problem of schema evolution and schema versioning support has been
extensively studied in relational and database papers. [Ro96] provides an excellent
survey on the main issues concerned. [Fr00] introduces an approach to schema
versioning, which considers a (conceptual) schema change as a (logical) schema
augmentation. In contrast to our approach, this semantic approach does not address
the change propagation problem, which concerns the effects of schema changes on the
underlying instances. For the change propagation problem, several solutions have
been proposed and implemented in real systems. In all cases, simple default
mechanisms can be used or user-supplied conversion functions must be defined for
non-trivial extant object updates. However, there are no approaches that treat data on
the web.
Annotation
We know of three major systems that intensively use knowledge markup in the
Semantic Web, viz. SHOE [HH00], Ontobroker [De99] and WebKB [PP99]. All three
of them rely on knowledge in HTML pages. They all started with providing manual
mark-up by editors. However, our experiences [Er00] have shown that text-editing
knowledge mark-up yields extremely poor results, viz. syntactic mistakes, improper
references, and all the problems sketched in the scenario section. The approaches
from this line of research that are closest to CREAM are the SHOE Knowledge
Annotator and the WebKB annotation tool. The SHOE Knowledge Annotator is a
Java program that allows users to mark-up webpages with the SHOE ontology. The
SHOE system [Lu97] defines additional tags that can be embedded in the body of
HTML pages. The SHOE Knowledge Annotator is rather a little helper (like our
earlier OntoPad [Fe99], [De99]) than a full-fledged annotation environment. WebKB
uses conceptual graphs for representing the semantic content of Web documents. It
embeds conceptual graph statements into HTML pages. Essentially they offer a Web-
based template like interface like knowledge acquisition frameworks described next.
6. Conclusion
Ontology used in an ontology-based KM system is related to the business strategy and
indirectly to the business environment. In the highly changed environment it is
obvious that an ontology as a domain backbone is also a matter of change. The
changes in the ontology have to be propagated to all ontology-based descriptions of
the knowledge sources in order to enable consistency of the searching process.
In this paper we have presented an approach for enabling consistency of the
descriptions of the knowledge sources in the case of the changes in the domain
ontology. The approach is based on our research in the ontology evolution and the
ontology-based annotation of the Web documents. The proposed method is
implemented in our semantic annotation framework so that efficient acquiring and
maintaining of the ontology-based metadata is supported.
The proposed approach has many benefits: automatic updating of the ontology-based
description of the content of the knowledge sources, new possibilities for searching
for the knowledge sources according to the author, date, format of the knowledge
sources, to name but a few. From the knowledge management system point of view
the proposed approach enables us to develop robust knowledge management solution,
which copes with the high-changeable business conditions.
Combining this approach with the ontology learning methods, which enable learning
ontologies from the knowledge sources, leads us in the some kind of self-organising
knowledge management systems.
Acknowledgements: We thank our colleagues and students at the Institute AIFB,
University of Karlsruhe, and FZI Research Center for Information Technologies at the
University Karlsruhe, without whom the research would not have been possible –
especially Steffen Staab and Alexander Mädche. Research for this paper was partially
financed by EU in the IST-2000-28293 project “OntoLogging” and by US Air Force
in the DARPA-DAML project “OntoAgent”.
7. Literature
[BL00] Berners-Lee, T.: XML 2000 – Semantic Web talk, 2000
http://www.w3.org/2000/Talks/1206-xml2k-tbl/slide10-0.html
[De99] Decker, S.; Erdmann, M.; Fensel, D. and Studer, R.: Ontobroker: Ontology Based
Access to Distributed and Semi-Structured Information. In Database Semantics: Semantic
Issues in Multimedia Systems, pages 351–369. Kluwer Academic Publisher, 1999.
[DPP00] Dzbor, M.; Paralic, J.; Paralic., M.: Knowledge Management in a Distributed
Organisation. In Advances in Networked Enterprises, Berlin, Germany, September 2000.
[Er00] Erdmann, M; Maedche, A.; Schnurr, H.-P. and Staab, S.; From manual to semi-
automatic semantic annotation. Proc. COLING 2000, Luxembourg, August 2000.
[Fe00] D.Fensel The role of Ontologies in Information Interchange. Proc. 2nd International
Scientific and Practical Conference on Programming UkrPROG'2000, Ukraine, 2000.
[Fe00a] Fensel, D.; Harmelen F; Klein, M; Akkermans, H; Broekstra, J.: On-To-Knowledge:
Ontology-based Tools for Knowledge Management: eBusiness and eWork, 2000
[Fe99] Fensel, D.; Angele, J.; Decker, S.; Erdmann, M.; Schnurr, H.-P.; Staab, S.; Studer, R.
and Witt, A.: On2broker: Semantic-based access to information sources at the www. In Proc. of
the World Conference on the WWW and Internet (WebNet 99), Honolulu, Hawaii, USA, 1999.
[Fi01] Firestone, J.: Knowledge Management Process Knowledge Management Process
Methodology: An Overview, Knowledge and innovation: Jornal of the KMCI, Volume 1, 2001
[FL96] Ferrandina, F.; Lautemann S.E.: An integrated approach to schema evolution for object
databases, OOIS 1996: 280-294
[Fr00] Franconi, E.; Grandi, F.; Mandreoli, F.: A semantic approach for schema evolution and
versioning in object-oriented databases. Proc. CL2000, 2000
[Gi99] Ginsburg, M.; Kambil, A.: Annotate: A Web-based Knowledge Management Support
System for Document Collections, Proc. of HICSS-32, 1999.
[Gr93] Gruber, T.: A translation approach to portable ontology specifications. Knowledge
Acquisition, 5:199–220, 1993
[KF01] Klein, M.; and Fensel, D.: Ontology versioning for the Semantic Web, Proc.
International Semantic Web Working Symposium (SWWS), USA, July 30 - August 1, 2001.
[Ha01] Handschuh, S.; Staab, S.; Mädche: A. CREAM — Creating relational metadata with a
component-based, ontology-driven annotation framework. ACM, K-CAP 2001, October, 2001
[Ha00] Hardless, C.; Lindgren, R.; Nulden, U. and Pessi, K.: The Evolution of knowledge
management system need to be managed, http://www.viktoria.informatik.gu.se/groups/
KnowledgeManagement/ Documents/kmman.pdf, 2000
[HH00] Heflin, J. and Hendler, J.: Searching the web with shoe. In Artificial Intelligence for
Web Search. Papers from the AAAI Workshop. WS-00-01, pages 35–40. AAAI Press, 2000.
[Lu97] Luke, S.; Spector, L.; Rager, D. and Hendler J.: Ontology-based Web Agents. In
Proceedings of First International Conference on Autonomous Agents, 1997.
[Ol99] Oliver D. E., Shahar Y., Musen M. A. and Shortliffe E. H.: Representation of change in
controlled medical terminologies. Artificial Intelligence in Medicine,15(1):53–76, 1999.
[PP99] Martin P. and Eklund P.: Embedding Knowledge inWeb Documents. Proc. 8th Int.
World Wide Web Conf. (WWW‘8), Toronto, 1999, pp. 1403–1419.
[Mc01] McElroy, M.: Second-Generation Knowledge Management, Macroinnovation
Associates, KMWorld 2001, November 2001, Santa Clara, Ca
[OL98] O’Leary, D.E.: Knowledge Management Systems: Converting and Connecting, IEEE
Intelligent Systems, 13, May/June, p.30-33.
[Ro96] Roddick, J.F.: A Survey of Schema Versioning Issues for Database Systems,
Information and Software Technology, 37(7):383-393, 1996.
[SM01] Staab, S. and Maedche, A.: Knowledge Portals - Ontologies at Work. AI Magazine,
21(2), Summer 2001.
[St01] Staab, S., H.-P. Schnurr, R. Studer, Y. Sure: Knowledge Processes and Ontologies. IEEE
Intelligent Systems. 16(1), January/February 2001. Special Issue on Knowledge Management.
[SS00] Staab, S.; Schurr, H.P.: Smart task support through proactive access to organizational
memory. Knowledge Based Systems 13(5): 251-260, 2000
[SSV02] Stojanovic, L.; Stojanovic N.; Volz R.: Migrating data-intensive Web Sites into the
Semantic Web. To appear: ACM Symposium on Applied Computing SAC, 2002.
[TB01] Tamma V.A.M. and Bench-Capon T.J.M; A conceptual model to facilitate knowledge
sharing in multi-agent systems. In Proc. of the OAS 2001. Montreal, May, 29, 2001, pp. 69-76.
... A similar approach was introduced by Stojanovic et al. (2002). Stojanovic et al. (2002) started to investigate how knowledge can be structured in a consistent way by using an ontology. ...
... A similar approach was introduced by Stojanovic et al. (2002). Stojanovic et al. (2002) started to investigate how knowledge can be structured in a consistent way by using an ontology. ...
... Their introduced approach automated the maintenance and updates of the knowledge base, but they did not introduce an ontology that could have been used for metadata (Stojanovic et al., 2002). ...
Thesis
Full-text available
Knowledge about data is often only available to certain experts which results in an inefficient search for data inside of companies. Recently, knowledge graphs have been recognised as tools to make knowledge retrieval easier in various knowledge domains. The goal of this thesis is to evaluate if a knowledge graph can improve the search for data inside of a company. In coorperation with a german bank the knowledge need of a data search was identified. The knowledge need consists of different knowledge domains, only when connecting those domains with each other, the respondend can find the data that he needs. The identified domains were designed as the data search ontology. The data search ontology was implemented as a knowledge graph prototype. By sucessfully querying the knowledge graph, it was validated that the designed ontology covers the identified knowledge need and can therefore be used to improve the search for enterprise data.
... With respect to the propagation of changes, it has only been considered in the past to related ontologies (e.g., [17], [25] and [9]), to ontology individuals (e.g., [26] and [27]), and to some extent to related applications ( [28]). For the propagation of changes to related ontologies, existing approaches consider only a central (main) copy of the ontology that is either replicated (e.g., [17]) or divided into several component ontologies (e.g., [9]) and where in general, changes are propagated only in one direction: from the main copy to its replicas. ...
... As we discussed in Section 2, there are several limitations in this area. In general, existing approaches (e.g., [17], [25]), [9], [26] and [27]) consider only the propagation to related ontologies and individuals. Besides, for the propagation of changes to related ontologies, these approaches consider only a central (main) copy of the ontology that is either replicated or divided into several component ontologies and, in general, changes are propagated only in one direction: from the main copy to its replicas. ...
... Pour cela, des ajustements importants sont probablement nécessaires. [Stojanovic et al., 2002a] [Stojanovic et al., 2002b] [Klein, 2004]. Ensuite, nous nous concentrons sur quelques approches qui traitent de problématiques particulières de gestion de changement telles que le processus d'évolution, la représentation d'un changement, la gestion de ses effets et de sa propagation vers les applications dépendantes. ...
... …………………………………………………………………………..…………572.4.2 Logiciels dédiés à l'évolution d'ontologie…………………………………………………………….582.4.2.1 Gestion des évolutions dans KAON[Stojanovic et al., 2002a],[Maedche et Staab, 2003] et [Gabel et al., 2004 ………………………………………………………………………………………………58 ...
Thesis
Full-text available
Dans un environnement dynamique, les ressources termino-ontologiques et les annotations sémantiques qu’elles permettent de construire doivent être modifiées régulièrement et en cohérence pour s’adapter à l’évolution du domaine sur lequel elles portent et des collections documentaires annotées. Notre contribution est d’associer l’évolution de la ressource termino-ontologique, les types de changements applicables (élémentaires ou complexes), ainsi que l’évolution des annotations sémantiques des documents. Cette contribution préserve la cohérence de la ressource termino-ontologique et conjointement celle des annotations sémantiques des collections documentaires. En support à un environnement d’annotation automatique de documents (TextViz) défini dans le cadre du projet DYNAMO qui prévoit plusieurs scénarios d’évolution, qui correspondent à des parcours différents, nous avons défini notre approche d’évolution EvOnto (Evolution d’Ontologie). L’originalité d’EvOnto est d’assurer un support à ces différents scénarios, tout en gérant une forte imbrication entre la ressource termino-ontologique et les annotations sémantiques. Il est destiné à un ontologue et le guide interactivement pour formuler une demande de changement, évaluer son impact (effets supplémentaires) sur la qualité de la ressource termino-ontologique et aussi sur les annotations sémantiques, et décider ensuite de leur mise en œuvre. Des informations sur l’utilisation de l’ontologie sont fournies à l’ontologue pour qu’il prenne l’initiative d’une évolution de la ressource termino-ontologique, en connaisse les conséquences, et les adapte pour minimiser les effets négatifs, les impacts non souhaitables ou les coûts correspondants sur la ressource elle-même et son utilisation dans des annotations.
... Ontology evolution consists in managing persistent ontology changes to cope with new requirements, and producing new versions. The modification of an ontology is handled by preserving its consistency, tracking and logging the change to provide mapping between subsequent versions (Stojanovic, Maedche, Motik, & Stojanovic, 2002a), and controlling the use of instances (Stojanovic, Stojanovic, & Handschuh, 2002b). ...
... Many researches have discussed the characteristics of an ontology evolution process (Klein, 2004;Stojanovic et al., 2002a;Stojanovic et al., 2002b) and several ontology evolution approaches have been proposed in the literature. Some focus on specific change management issues like capturing change requirements (Stojanovic, Stojanovic, Gonzalez, & Studer, 2003a;Cimiano & Völker, 2005;Bloehdorn, Haase, Sure, & Voelker, 2006), change detection and version logging (Klein, Fensel, Kiryakov, & Ognyanov, 2002a;Noy, Kunnatur, Klein, & Musen, 2004;Plessers & De Troyer, 2005;Eder & Wiggisser, 2007), formal change specification (Stojanovic, Stojanovic, & Volz, 2002c;Klein, 2004;Plessers De Troyer, & Casteleyn, 2007), change implementation (Stojanovic, Maedche, Stojanovic, & Studer, 2003b;Stojanovic, 2004;Flouris, 2006), consistency maintenance (Stojanovic, 2004 ;Haase & Stojanovic, 2005;Haase & Völker, 2005;Plessers & De Troyer, 2006), ontology versioning (Klein & Fensel, 2001;Klein et al., 2002a;Klein, 2004), and others propose a more or less global evolution process including change impact analysis and resolution as well as change propagation to dependant artifacts (objects, ontologies and applications referenced by the ontology) (Stojanovic, 2004;Klein, 2004;Bloehdorn et al., 2006). ...
Article
Ontologies evolve continuously throughout their lifecycle to respond to different change requirements. Several problems emanate from ontology evolution: capturing change requirements, change representation, change impact analysis and resolution, change validation, change traceability, change propagation to dependant artifacts, versioning, etc. The purpose of this chapter is to gather research and current developments to manage ontology evolution. The authors highlight ontology evolution issues and present a state-of-the-art of ontology evolution approach by describing issues raised and the ontology model considered (ontology representation language), and also the ontology engineering tools supporting ontology evolution and maintenance. Furthermore, they sum up the state-of-the-art review by a comparative study based on general characteristics, evolution functionalities supported, and specificities of the existing ontology evolution approaches. At the end of the chapter, the authors discuss future and emerging trends.
... Ontology evolution phases and their issues have been described by the different researchers such as capturing change requirements described by [14,15], change detection and version logging described by [16], formal change specification described by [12], change implementation and consistency maintenance described by [13]. ...
Article
Full-text available
Ontology‐based knowledge‐driven activity recognition (AR) models play a vital role in realm of Internet of Things (IoTs). However, these models suffer the shortcomings of static nature, inability of self‐evolution, and lack of adaptivity. Also, AR models cannot be made comprehensive enough to cater all the activities and smart home inhabitants may not be restricted to only those activities contained in AR model. So, AR models may not rightly recognise or infer new activities. Here, a framework has been proposed for dynamically capturing the new knowledge from activity patterns to evolve behavioural changes in AR model (i.e. ontology based model). This ontology‐based framework adapts by learning the specialised and extended activities from existing user‐performed activity patterns. Moreover, it can identify new activity patterns previously unknown in AR model, adapt the new properties in existing activity models and enrich ontology model by capturing change representation to enrich ontology model. The proposed framework has been evaluated comprehensively over the metrics of accuracy, statistical heuristics, and Kappa coefficient. A well‐known dataset named DAMSH has been used for having an empirical insight into the effectiveness of proposed framework that shows a significant level of accuracy for AR models.
... Knowledge artifacts link the portfolio view with the activities of employees. For being able to describe the knowledge artifacts in a structured way, a metamodel was established and is partially presented in Fig. 3. Defining the knowledge metamodel and, thus, specifying the structure of knowledge artifacts, is an important prerequisite for integrating different knowledge sources [5]. ...
Conference Paper
Knowledge is one of the most important resources today and, thus, companies in general and SMEs in particular need effective and efficient knowledge management solutions. In this paper, activities in the project MACKMA, which aims at implementing a knowledge management system tailored to the needs of SMEs, are introduced. Central findings include methods for establishing the product-service-portfolio of a company, a metamodel for knowledge artifacts and an accompanying incentive system for increasing knowledge management system usage.
... In addition, we will enhance the change categories and interesting changes with the feedback and group single changes to more human-understandable composite changes similarly to the work of Stojanovic et al. [6] ...
Article
Full-text available
We consider ontology evolution in a system of light-weight Linked Data ontologies, aligned with each other to form a larger ontology system. When one ontology changes, the human editor must keep track of the actual changes and of the modifications needed in the related ontologies in order to keep the system consistent. This paper presents an analysis tool MUTU, by which such changes and their potential effects on other ontologies can be found. Such an analysis is useful for the ontology editors for understanding the differences between ontology versions, and for updating linked ontologies when changes occurred in other components of an ontology system.
... In the literature, several approaches have been taken into account to address ontology evolution: the ontology evolution approach and the versioning ontology approach. The evolution of the ontology is defined in (Stojanovic et al., 2002) as the ability to update the existing ontology following the emergence of new needs and maintain its consistency and coherence. It maintains a unique ontology (the last one). ...
Chapter
Ontologies are like database schema and schema versioning in temporal databases can be useful in order to propose an approach for ontology versioning. In fact, we can benefit from principles and tools we previously defined for schema versioning in such databases, in order to ensure an efficient management of versions in ontological databases. We are interested in developing an ontology versioning system to express, apply and implement changes on the ontology. The adopted ontology versioning approach is based on three steps: evolution changes, ontology coherence and versioning management. Our goal is to assist users in expressing evolution requirements, observing their consequences on the ontology and comparing ontology versions.
... ˆ Change representation: in order to be correctly implemented, we have to represent these causal changes formally, explicitly and in a suitable format. In the context of SemCaDo algorithm, we only handle elementary changes [100] (i.e. restricted to adding semantic causal relations) that cannot be decomposed into simpler ones. ...
Article
Full-text available
With the rising need to reuse the existing domain knowledge when learning causal Bayesian networks, the ontologies can supply valuable semantic information to de ne explicit cause-to-e ect relationships and make further interesting discoveries with the minimum expected cost and e ort. This thesis studies the crossing-over between causal Bayesian networks and ontologies, establishes the main correspondences between their elements and develops a cyclic approach in which we make use of the two formalisms in an interchangeable way. The rst direction involves the integration of semantic knowledge contained in the domain ontologies to anticipate the optimal choice of experimentations via a serendipitous causal discovery strategy. The semantic knowledge may contain some causal relations in addition to the strict hierarchical structure. So instead of repeating the e orts that have already been spent by the ontology developers and curators, we can reuse these causal relations by integrating them as prior knowledge when applying existing structure learning algorithms to induce partially directed causal graphs from pure observational data. To complete the full orientation of the causal network, we need to perform active interventions on the system under study. We therefore present a serendipitous decision-making strategy based on semantic distance calculus to guide the causal discovery process to investigate unexplored areas and conduct more informative experiments. The idea mainly arises from the fact that the semantically related concepts are generally the most extensively studied ones. For this purpose, we propose to supply issues for insight by favoring the experimentation on the more distant concepts according to the ontology subsumption hierarchy. The second complementary direction concerns an enrichment process by which it will be possible to reuse these causal discoveries, support the evolving character of the semantic background and make an ontology evolution. Extensive experimentations are conducted using the well-known Saccharomyces cerevisiae cell cycle microarray data and the Gene Ontology to show the merits of the SemcaDo approach in the biological eld where microarray gene expression experiments are usually very expensive to perform, complex and time consuming.
... • Change representation: in order to be correctly implemented, we have to represent these causal changes formally, explicitly and in a suitable format. In the context of SemCaDo algorithm, we only handle elementary changes [57] (i.e. restricted to adding semantic causal relations) that cannot be decomposed into simpler ones. ...
Article
Full-text available
Within the last years, probabilistic causality has become a very active research topic in artificial intelligence and statistics communities. Due to its high impact in various applications involving reasoning tasks, machine learning researchers have proposed a number of techniques to learn Causal Bayesian Networks. Within the existing works in this direction, few studies have explicitly considered the role that decisional guidance might play to alternate between observational and experimental data processing. In this paper, we go further by introducing a serendipitous strategy to elucidate semantic background knowledge provided by the domain ontology to learn the causal structure of Bayesian Networks. We also complement our contribution with an enrichment process by which it will be possible to reuse these causal discoveries, support the evolving character of the semantic background and make an ontology evolution. Finally, the proposed method will be validated through simulations and real data analysis.
Article
Full-text available
Today many organizations are reliant on the knowledge and competence of individual organizational members. Information systems to support knowledge management (KM) are therefore considered to be vital tools in order to achieve competitive advantage. In this paper, we report the results from a field research study of such systems in a knowledge-intensive, fast-growing and dynamic organization. The case illustrates that evolution, which refers to the process by which organizations and their information systems change over time, needs to be managed since it can result in KM-systems failures. We characterize the mainstream KM research literature in relation to managing the risk of KM-systems failures, and outline that management of KM-systems' evolution is a dimension that has not been addressed so far. With these empirical and theoretical results as a basis, we argue that more attention must be given to managing the evolution of KM-systems.
Article
Full-text available
Schema versioning is one of a number of related areas dealing with the same general problem—that of using multiple heterogeneous schemata for various database related tasks. In particular, schema versioning, and its weaker companion, schema evolution, deal with the need to retain current data and software system functionality in the face of changing database structure. Schema versioning and schema evolution offer a solution to the problem by enabling intelligent handling of any temporal mismatch between data and data structure. This survey discusses the modelling, architectural and query language issues relating to the support of evolving schemata in database systems. An indication of the future directions of schema versioning research is also given.
Conference Paper
Full-text available
In this paper a semantic approach for the specification and the manage- ment of databases with evolving schemata is introduced. It is shown how a general object-oriented model for schema versioning and evolution can be formalized; how the semantics of schema change operations can be defined; how interesting reasoning tasks can be supported, based on an encoding in description logics.
Conference Paper
Full-text available
On2broker provides brokering services to improve access to heterogeneous, distributed and semistructured information sources as they are presented in the World Wide Web. It relies on the use of ontologies to make explicit the semantics of web pages. In the paper we will discuss the general architecture and main components of On2broker and provide some application scenarios.
Article
Full-text available
this article, we present an approach for ontology -based KM that includes a suite of ontologybased tools as well as a methodology for developing ontology-based KM systems. Our approach, shown in Figure 1, builds on the distinction between knowledge process (handling knowledge items) and knowledge metaprocess (introducing and maintaining KM systems). Ontologies constitute the glue that binds knowledge subprocesses together. Ontologies open the way to move from a document-oriented view of KM to a content-oriented view, where knowledge items are interlinked, combined, and used. The method for developing KM systems that we outline in this article (that is, the knowledge metaprocess) extends and improves the CommonKADS method 3 by introducing specific guidelines for developing and maintaining ontologies. Our approach shows that you can clearly identify and handle different subprocesses that drive the development and use of
Article
We describe an approach towards integrating the semantics of semi-structured documents with task-support for (weakly structured) business processes and proactive inferencing capabilities of a desk support agent. The mechanism of our Proactive Inferencing Agent is motivated by the requirements posed in (weakly structured) business processes performed by a typical knowledge worker and by experiences we have made from a first trial with a Reactive Agent Support scheme.Our reactive scheme is an innovative approach for smart task support that links knowledge from an organizational memory to business tasks. The scheme is extended to include proactive inferencing capabilities in order to improve user-friendliness and to facilitate modeling of actual agent support. In particular, the improved scheme copes with varying precision of knowledge found in the organizational memory and it reasons proactively about what might be interesting to you and what might be due in your next step.