=Paper=
{{Paper
|id=Vol-2267/447-452-paper-85
|storemode=property
|title=Semantic information management: the approach to semantic assets development lifecycle
|pdfUrl=https://ceur-ws.org/Vol-2267/447-452-paper-85.pdf
|volume=Vol-2267
|authors=Yury Akatkin,Elena Yasinovskaya,Mikhail Bich
}}
==Semantic information management: the approach to semantic assets development lifecycle==
https://ceur-ws.org/Vol-2267/447-452-paper-85.pdf
Proceedings of the VIII International Conference "Distributed Computing and Grid-technologies in Science and
Education" (GRID 2018), Dubna, Moscow region, Russia, September 10 - 14, 2018
SEMANTIC INFORMATION MANAGEMENT: THE
APPROACH TO SEMANTIC ASSETS DEVELOPMENT
LIFECYCLE
Yu. Akatkin a, E. Yasinovskaya b, M. Bich c
Plekhanov Russian University of Economics, 36, Stremyanny lane Moscow, Russia 117997
E-mail: a uakatkin@semanticpro.org, b elena@semanticpro.org, c misha@semanticpro.org
The application of semantic integration methods faces challenges arising at collaboration between IT-
specialists and domain experts during the model building stage. These challenges can affect the
correctness of domain formalization as well as the whole outcome of the integration in distributed
information systems. To overcome the lack of semantic interoperability we suggest the creation of a
collaborative platform which provides the (re)use of semantic assets (SA) for further semantic
integration. The analysis of the limitations existing in SA management standards leads the authors to
propose the collaborative approach, based on an extended lifecycle of semantic assets. The authors
consider the implementation of the platform based on the Asset Description Metadata schema
extension to be a rational option.
Keywords: Semantic interoperability, semantic information management, lifecycle of semantic assets,
semantic integration, distributed information systems, ADMS, model driven architecture
© 2018 Yury Akatkin, Elena Yasinovskaya, Mikhail Bich
447
Proceedings of the VIII International Conference "Distributed Computing and Grid-technologies in Science and
Education" (GRID 2018), Dubna, Moscow region, Russia, September 10 - 14, 2018
1. Introduction
Lots of scientific papers describe the application of semantic methods in the integration of data
from heterogeneous sources [1, 2, 4, 9, 10, 12, 14, 17]. Currently distributed systems mostly support
interoperability on technical and organizational levels as a rule. Although, semantic (information)
interoperability becomes essential for successful integration. The effective implementation of semantic
methods for the development of e-Government in Europe and the United States1 over the last 15 years
[6, 12, 14] has proved this statement in practice.
The ability to understand the meaning of data determined by the provider becomes extremely
important. It improves the growth of semantic interoperability significance in heterogeneous
environment with constantly changing number of participants. Data-centric paradigm is now the key
enabler for the development of disruptive technologies in a new digital world [7, 16].
The use of semantic data models [6] (e.g. glossaries, dictionaries, taxonomies, thesauri and
ontologies) – hereinafter referred to as semantic assets (SA) [13] – is the basis for semantic
integration. SA enable semantic interoperability of distributed information systems (IS) and serve for
data collection, search, analysis and data visualization performed by using semantic properties.
As a rule, groups of domain experts construct SA throughout the process of information
system development or modeling of a particular domain. It is important to transfer the domain
knowledge from paper documents into machine-readable formats (ontologies, thesauri, and glossaries)
and to provide its dissemination outside of a specific information system. It brings the “understanding”
of data to the information exchange and simplifies further SA (re)use by other IS interacting in a
heterogeneous environment. At the same time, already developed SA are often insufficient for the
modeling of information exchange, therefore, experts come across the task to build a more detailed
semantic exchange model based on SA applied to the domain of interacting IS.
Existing ontology-based approaches for semantic interoperability have not been sufficiently
effective because “there is no systematic methodology to follow, no concert methodology for building
ontologies and all existing ontology-based not able to reconcile all types of semantic conflicts”.
Ontology Summit supports this experts’ concern and emphasizes that “in practice, however, Semantic
Interoperability is difficult to achieve” [11].
We consider it necessary to join the efforts of IT-specialists and various domain experts
providing the ability of information systems to interact on a semantic level. Additionally, their
cooperation should help to solve the problem of cross disciplinary misunderstanding which results in
multiple revisions and unsatisfactory results. To overcome the challenges shown above and to simplify
the application of semantic integration methods at the stage of semantic asset development we suggest
setting an expert-oriented, common methodology for SA Management and support it with appropriate
tools.
2. Semantic Assets Management
Semantic Web expansion have caused the development of many semantic assets that have
become standards and recommendations, which de facto or de jure define various components or data
schemes (e.g. Dublin Core, FOAF, VOAF, SKOS, vCard, etc.). The need to use solutions supporting
accessibility and (re)use of semantic assets, considering special aspects of localization, has become
obvious for the accumulation and spread of knowledge encapsulated in SA. Within the general
guidelines some projects (e.g. JOINUP [12, 13]) stipulate the methodology for semantic assets’
management. However, existing SA cataloguing platforms do not provide a complete lifecycle
especially at the stages of SA development and modernization.
Within the framework of the European Interoperability program ISA [6] in 2011-2012 the
ADMS Working Group [12] developed the Asset Description Metadata Schema (ADMS)2 to collect,
search and study the compatibility of semantic assets. In 2013, the extended ADMS (ADMS-AP 1.0)
specification was developed. This profile was focused on semantic interoperability based on the
1
NIEM 2017. National Information Exchange Model, https://www.niem.gov
2
W3C 2013. Asset Description Metadata Schema (ADMS). W3C. http://www.w3.org/TR/vocab-adms/
448
Proceedings of the VIII International Conference "Distributed Computing and Grid-technologies in Science and
Education" (GRID 2018), Dubna, Moscow region, Russia, September 10 - 14, 2018
unification of SA descriptions. In 2016, following the use experience of ADMS-AP 1.0, the revision
of the specification resulted in the adoption of the version of ADMS-AP 2.0 [5]. ADMS and,
accordingly, JOINUP platform, describe SA’s lifecycle using four statuses: Completed,
UnderDevelopment, Deprecated, and Withdrawn. We consider the use of ADMS for SA cataloguing,
as well as, for storing and publishing their descriptions. But we suppose in practice this lifecycle
includes some additional steps and should be extended for the following stages of SA development:
(1) experts’ collaboration, (2) validation, (3) assessment and quality evaluation of represented
semantic assets.
We have also studied W3C recommendations [15], and ISO/IEC 11179 [8] and the conclusion
is both support the lifecycle but manage the entity as a single document. The adoption of these
approaches is reasonable only for SA management without considering the constituent elements.
However, at the stage of SA development/modernization the contents of semantic asset, i.e. its
elements and their properties, play the main role.
Therefore, we propose to combine the lifecycle management of semantic models as a single
asset (e.g. ADMS repository records) with the change management of their contents. The reason for
this consolidation is the ability of SA contents to split into parts (branches, sections, sets and
elements), each of which could follow its own specific workflow and should be controlled, reviewed
and assessed during the stage of development.
We think it reasonable to use a widespread and recognized by web community W3C lifecycle
to describe the development stages of SA (ADMS status “In Development”). It reflects the features of
an open SA catalogue such as attracting a wide range of experts, a variety of SA domains, use of SA in
Semantic web and exchange of information using web-based technologies.
3. Semantic Assets Development Lifecycle
To bridge the gap associated with the lack of understanding between domain and IT experts in
the development of information systems and their interaction, a collaborative semantic integration
platform [3]. Working on the SA lifecycle management methodology during our R&D initiative to
create the Center for Semantic Integration (CSI) at the Plekhanov Russian University of Economics
we revealed and covered the following scenarios:
• Domain experts develop semantic assets in scope and detail necessary for IT specialists
during SA implementation in information systems.
• Domain experts establish the correspondence between different SA elements to harmonize
their content.
• IT specialists ensure the consistency of the developed semantic models with the description
of the subject area, generally accepted standards and recommendations regarding the
composition and contents.
• Domain experts review the models developed by IT specialists to assess semantic
completeness and consistency.
Basing on these scenarios we set the primary task to support both (1) the development of
semantic assets and (2) the harmonization of SA contents. We register existing semantic assets in the
catalogue, based on ADMS. After expert review they can be loaded from external resources for further
(re)use (e.g. for localization). The level of expert review is selected depending on the level interest: (1)
validation among a working group, (2) public expert assessment by the community, (3) an independent
review conducted by domain experts.
Throughout the collaboration of working group participants during the SA development,
various stages can occur (see Figure 1), which can be described by W3C workflow statuses.
449
Proceedings of the VIII International Conference "Distributed Computing and Grid-technologies in Science and
Education" (GRID 2018), Dubna, Moscow region, Russia, September 10 - 14, 2018
Figure 1. Lifecycle of semantic assets
We offer to arrange the process of SA development by assigning business tasks and
controlling their performance. In Figure 1, the solid lines indicate the possible transitions between the
lifecycle states or the development stages; and broken lines show the connections between the states,
stages, and events that occur during the SA lifecycle. Each stage of SA development has a set of
business tasks available for assignment (see Table 1).
Table 1. Stages of SA development stages and business tasks
Business tasks for SA
SA development stages Comments
development stages
Filling the main content of the
asset (loading SA contents,
SA Loading
creation of new SA elements,
SA Creation
Working Draft (WD) translation, etc.) and refining the
SA Translation
description of SA (completion
SA expert assessment
of fields, classification,
connection to other assets).
Main expert evaluation and
Candidate for recommendation SA Modification
making small changes, e.g.
(CR) SA expert assessment
spelling improvement.
Decision that the SA can be
Proposed recommendation (PR) SA expert assessment recommended for
implementation.
Recommendation (REC)
450
Proceedings of the VIII International Conference "Distributed Computing and Grid-technologies in Science and
Education" (GRID 2018), Dubna, Moscow region, Russia, September 10 - 14, 2018
Implementation of methods and tools for the reuse of semantic assets including constituent
elements imposes restrictions on SA deletion/dismissal. In addition to the usual notification, informing
SA users about the changes, the methodology should cover the events associated with the SA
withdrawal (ADMS-AP 2.0 lifecycle status “Deprecated” or “Withdrawn”). The existing catalogues
serve only for SA collection and provide several versions of a semantic asset existing in parallel.
However, it is necessary to preserve the knowledge about the changes in semantics and the link
between “old” and “new” semantic assets should be established.
To support the lifecycle of semantic assets we have developed the extension to ADMS-AP 2.0,
named ADMS-WF 1.0. It provides compatibility and reuse of SA by domain experts and IT specialists
at the stage of development. For the implementation of the usage scenarios represented above we offer
to extend ADMS-AP 2.0 classes and add:
additional properties to ADMS “Asset”, dcat:Dataset class for the conduct of SA
versioning, access to SA contents together with the connection to business tasks and
events.
additional property to ADMS “Asset Distribution”, dcat: Distribution class for the
integration with external design tools.
Besides that, it is necessary to add:
“Event”, dct:Event class for the storage and distribution of events, occurring during
SA development.
“Change Request”, cm:ChangeRequest class in order to fix business tasks connected
with SA as well as their workflow.
Special semantic models such as controlled vocabularies of events, business tasks types,
design tools and SA lifecycle states determine available property values. ADMS-WF 1.0 supports
backward compatibility of SA catalogues, implemented in CSI collaboration platform, with
repositories, correspondent to ADMS-AP 2.0.
4. Conclusion
This article presents the approaches to the achievement of semantic interoperability in
heterogeneous information systems environment. To provide an unambiguous, meaningful
interpretation of data by all the participants of information sharing we keep to the following principles:
(1) consolidation and reuse of semantic assets, and (2) the collaboration of domain experts and IT
specialists working with semantic assets.
We consider the methodology for such collaboration mainly based on the reuse existing
methods and Web standards. Due to the analysis of limitations inherent to known SA management
standards we reveal the necessity to extend SA lifecycle and to support the expert workflow at the
stage of SA development. Following the usage scenarios determined during our R&D at the CSI
project we suggest combining lifecycle of SA (including the elements), SA development stages and
lifecycle of business tasks.
To implement the proposed approaches, we have developed an extension to ADMS-AP 2.0 –
ADMS-WF providing backward compatibility of the collaborative semantic integration platform with
ADMS-compliant catalogues together. It also serves for information sharing and dissemination of SA
built or modified in the process of experts’ collaboration for the semantic integration. Companies
owning semantic assets, experts and IT specialists can use the ADMS-WF 1.0 profile for cataloging,
as well as for supporting the lifecycle of semantic assets during harmonization, developing new or
finalizing existing versions.
In addition, experts can use the ADMS-WF 1.0 to support the lifecycle of SA during the
process of semantic assets discussion and assessment. IT specialists can use this profile to support the
lifecycle of SA when validating the correspondence of data schemas (metadata) to domain models or
enriching data schemas with the semantic information needed to support interoperability of
heterogeneous information systems.
451
Proceedings of the VIII International Conference "Distributed Computing and Grid-technologies in Science and
Education" (GRID 2018), Dubna, Moscow region, Russia, September 10 - 14, 2018
References
[1] Akatkin Yu.M., Yasinovskaya E.D., Bich M.G. 2016. Management and (re)use of semantic assets
for information sharing. DAMDID / RCDL'2016, 235–242, http://elibrary.ru/item.asp?id=27401010
[2] Akatkin Yu.M., Yasinovskaya E.D, Bich M.G., Shilin A.V. 2017. Application of semantic
integration methods on the example of web services information exchange arrangement. Information
Resources of Russia №2 (156) – 2017, P.13-19
[3] Akatkin Yu.M., E.D. Yasinovskaya, Bich M.G 2016. Methods of Semantic Integration in
Distributed Information Systems: Challenges of Application. 1787 (2016), 80–84. http://ceur-
ws.org/Vol-1787
[4] Date C. J. 2004. An Introduction to Database Systems (8 ed.). Pearson Education Inc., Pearson
Education Inc. P. 1024. ISBN 0-321-18956-6. P247
[5] European Commission 2016. ADMS Application Profile for solutions on Joinup Version 2.01.
European Commission. https://joinup.ec.europa.eu/node/150484/
[6] European Commission. 2010. Annex 2 to the Communication from the Commission to the
European Parliament, the Council, the European Economic and Social Committee and the Committee
of Regions «Towards interoperability for European public services», COM (2010) 744 final.
[7] Fernández, A. 2018. Putting value at the core of your business: Data-driven vs Data-centric,
Stratio, http://www.stratio.com/blog/datadriven-versus-datacentric/
[8] International Organization for Standardization. International Standard ISO/IEC 11179-6:2015,
Information technology - Metadata registries (MDR) - Part 6: Registration. International Organization
for Standardization
[9] Madnick, S., Gannon, T., Zhu, H., Siegel, M., Moulton, A., Sabbouh, M. 2009. Framework for the
Analysis of the Adaptability, Extensibility, and Scalability of Semantic Information Integration and
the Context Mediation Approach. Massachusetts Institute of Technology Cambridge. MA. USA
http://web.mit.edu/smadnick/www/wp/2009-02.pdf
[10] Madnick, S.E., & Zhu, H. 2006. Improving data quality through effective use of data semantics,
Data and Knowledge Engineering, 59(2). P.460-475
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.89.3747&rep=rep1&type=pdf
[11] Ontology Summit. Ontologies within Semantic Interoperability Ecosystems. In Ontology Summit
2016 Communique. http://ontologforum.org/index.php/ OntologySummit2016/Communique
[12] Vassilios Peristeras, Gofran Shukair, Nikolaos Loutas. 2012. Towards semantically interoperable
metadata repositories: The Asset Description Metadata Schema. Computers in Industry 64 (2013)
(2012), 10–18. https://www.sciencedirect.com/science/article/pii/S0166361512001406
[13] Vassilios Peristeras. 2011. Federation of Semantic Assets Repositories.
https://joinup.ec.europa.eu/sites/default/files/08/5e/bb/federation-of-semantic-assets-repositories.pdf
[14] Vassilios Peristeras. 2013. Semantic Standards: Preventing Waste in the Information Industry.
IEEE Intelligent Systems, 72–75. https://doi.org/10.1109/MIS.2013.115
[15] W3C. W3C Consortium Process Document. W3C. http://www.w3.org/Consortium/Process/
[16] Wahlin, R.: Data-centric architectural best practices: Using DDS to integrate real-world
distributed systems, Military Embedded Systems, http://mil-embedded.com/articles/data-centric-real-
world-distributed-systems/
[17] Walaa, S. Ismail, Mona M. Nasr, Torky I. Sultan, Ayman E. Khedr. 2013. Semantic Conflicts
Reconciliation as a Viable Solution for Semantic Heterogeneity Problems. (IJACSA) International
Journal of Advanced Computer Science and Applications. Vol. 4. No.4
https://pdfs.semanticscholar.org/22bb/9c1c27049fce16d1546c54d51e0ede2854d8.pdf
452