Posts Tagged ‘ontology evolution’

Moulding USDL in SBVR using Business Semantics Glossary: Part 1

July 2, 2011

I truly believe in co-creation. For example, we have our Collibra software and methods regularly scrutinized by numerous master students from both technical as well as more business-oriented computer science programmes in universities across Europe.

At VU University, for example, in the context of my Business Semantics Management master class, 21 MSc students playing the role of steward formed the Amsterdam Service Modelling Community with one common purpose: building an SBVR version of the USDL service description language. There were two additional members invited playing the role of observer: Carlos Pedrinaci (representing the USDL W3C incubator group) and Ivan Razo-Zapata (our PhD student at VU working on dynamic service market place composition). Finally there was me playing the role of administrator, making a total of 24 members.

The Amsterdam Service Modelling community has 24 members (21 playing the role of "steward"; 2 as "observer"; 1 as "administrator") and is subdivided in 5 speech communities. (Note: this is an older release of the Business Semantics Glossary software)

The figure below depicts the Business Semantics Management (BSM) methodology that is established by two operational cycles (reconciliation and application) each grouping a number of modeling activities. For a summary go here and for more details see my dissertation.

Business semantics management is established by two operational cycles each grouping a number of modeling activities.

The experiment extended over a total period of 4 weeks; hence we limited ourselves to the first 4 steps of semantic reconciliation only: scope, create, refine, articulate. In September we plan to repeat this experiment  over a period of 8 weeks where we will have time to do one full cycle of BSM. Later I will also blog about similar experiments we conduct at VUB University of Brussels.

Community-driven Approach

The Amsterdam Service Modelling Community (ASMC) is modelled (in SBVR) as a semantic community. SBVR takes into account the existence of multiple perspectives on how to represent concepts (by means of vocabularies).

  • A semantic community is a group of stakeholders having a body of shared meanings. Stakeholders are people representing an organisation or a business unit.
  • A body of shared meanings is a unifying and shared understanding (perception) of the business concepts in a particular domain.  Concepts are identified by a URI.
  • A speech community is a sub-community of a semantic community having a shared set of vocabularies to refer to the body of shared meanings. A speech community groups stakeholders and vocabularies from a particular natural language in a multi-lingual community, or from a certain technical jargon.
  • A vocabulary is a set of terms and fact types (called vocabulary entries) primarily drawn from a single language to express concepts within a body of shared meanings.

Within the ASMC community, the 21 students grouped in 5 speech communities each focusing on a specific part of the USDL framework. In SBVR, speech communities are part of one semantic community and each manage their own set of vocabularies to refer to this body of shared meanings. This allows for different representations of the same business concepts.

The navigator shows (from left to right) the structure of communities and their vocabularies.

Scoping the Semantic Reconciliation Cycle

The module-based decomposition of USDL depicted below makes it easy for teams to scope. However, they all had to start from the Service and Pricing module so we could observe divergence in definitions as well, an important step in the ontology evolution process (see the Perspective Rendering principle of my PhD on BSM).

The module-based USDL framework allows for clear scoping among speech communities (by courtesy of http://www.internet-of-services.com/).

Create, Refine, Articulate

Below is a screenshot of the term “Service” in the Pricing and Participant vocabulary in development by the “VAAF” speech community team. The steward (indicated on the top-right) “Vlant” is responsible for selecting the right stakeholders (bottom-right) among his fellow members and engage them into the reconciliation of the term.

A term can be defined using different kinds of attributes, going from (business-oriented) descriptions and definitions to more (formal) fact types and business rules.  Currently the level of articulation is below threshold (37.5%) incentivizing the steward and stakeholders to elaborate more.

Term "Service" in the Pricing and Participant vocabulary in development by the "VAAF" speech community team.

Next time we will talk about vocabulary statistics and workflows in the software. Workflows practically implement the orchestration of reconciliation tasks to members according to their roles and responsibilities.

Advertisements

Social Performance in Collaborative Business Semantics Management

September 9, 2009

The “living” ontologies that will furnish the Semantic Web are lacking. The problem is that in ontology engineering practice, the underlying methodological and organisational principles to involve the community are mostly ignored. Each of the involved activities in the community-based ontology evolution methodology require certain skills and tools which domain experts usually lack. Finding a social arrangement of roles and responsibilities that must supervise the consistent implementation of methods and tools is a wicked problem. Based on three technology-independent problem dimensions of ontology construction, we propose a set of social performance indicators (SPIs) to bring insights in the social arrangement evolving the ontology, and how it should be adapted to the changing needs of the community. We illustrate the SPIs on data from a realistic experiment in the domain of competency-centric HRM.

Actions grouped per part of the ontology over time.

Actions grouped per part of the ontology over time.

The illustration here is a sneak preview of what’s to come. It is the analysis of an SPI that observes the balance between the human resources spent on the respective parts (computational\formal vs. substantial\informal parts) of the representation of individual concept types through time. This may indicate the need to adapt the social arrangement accordingly.  The actions are grouped per part of the ontology: G0 for the discussion part; G1 for the formal part; and G2 for the informal part. G3, 4 and 5 resp. for creating, deleting and moving concept pages. The graph shows three moments (i.e., 3/26; 4/2; and 4/23) where all groups peak. These moments indicate (i) an intermediary deadline for a new ontology version to be accepted, and (ii) and consequently a point where the domain is rescoped for another iteration of the ontology evolution cycle, resulting in a temporarily higher production. The initial scoping peak is the largest, while the following two peaks become gradually smaller. This indicates the ontology reaches a fixpoint as the final deadline approaches, as more concepts covering the domain become mature. There are two isolated peaks of actions on the formal parts in the second iteration: 29 actions on 2009-04-09 and 22 on 2009-04-16. This shift of balance between formal and informal actions is the result of a general request by the core domain expert to spent more resources on formalisation of core concept types.

The full experiment will be presented at and published by the International Semantic Web Conference (ISWC) 2009.

Full article: De Leenheer, P., Debruyne, C., Peeter, J. (2009) Towards Social Performance Indicators for Community-based Ontology Evolution. In Proc. of ISWC Workshop on Collaborative Construction, Management and Linking of Structured Knowledge (CK2008)

The role of Semantics in the future Internet

May 12, 2009

Here is an easy movie that presents the future Internet.  In my article  that was published last year at the Future Internet Symposium 2008 held in Vienna, Austria I discussed to role of business semantics management in here, and the need for a kind of DNS of information consisting of data services compiled from and orchestrated by semantic patterns . See my blog about this here

International Workshop on Community-Based Evolution of Knowledge-Intensive Systems

April 27, 2009

Vilamoura, Portugal: November 1-6, 2009

Proceedings will be published by Springer Verlag

Workshop Theme

COMBEK seeks to address the need for research that explores and embraces the novel, difficult but crucial issue of adapting knowledge resources to their user communities, and vice versa, as a fundamental property of knowledge-intensive internet systems. Through a deep understanding of the real-time, community-based, evolution of so-called ontologies, a knowledge-intensive system can be made operationally relevant and sustainable over long periods of time.

Basic principle of knowledge-intensive community systems is the co-evolution of (A) social interactions enabled by the community’s design; (B) the open applications development designated to support them; and (C) the semantic patterns to enable semantic interoperability between these applications.

By addressing the notion of “community” in this way, COMBEK hopes to innovate the science of ontology engineering and unlock the expected (and unavoidable) paradigm shift in knowledge-based and community-driven systems.  Such a paradigm would affect knowledge sharing and communication across diverse communities in business, industry, and society.  We are further convinced that being a part of the OnTheMove  conferences will turn a spotlight on the scientific issues addressed in COMBEK, making them visible and attractive to industry.

Workshop Goals

COMBEK is ready to transcend the current, narrow “ontology engineering” view on the change management of knowledge structures that is at the heart today’s knowledge-intensive systems. We will consider stakeholder communities as integral factors in the continuous evolution of the knowledge-intensive systems in which they collaborate. By bringing together researchers from different domains, COMBEK aims to advance research on a very broad spectrum of needs, opportunities, and solutions. COMBEK will be a forum for the discussion of next-generation knowledge-intensive systems and radically new approaches in knowledge evolution.

Topics of Interest

COMBEK goes wider than current practice by accepting explorations of new and alternative approaches from multiple relevant disciplines, including, but not limited to:

  • collaborative knowledge engineering
  • pragmatic web / pragmatic semantic unification
  • community-driven knowledge acquisition and sharing
  • ontology negotiation and argumentation
  • community-driven ontology evolution management
  • change impact analysis
  • lexical resources for ontology representation and disambiguation
  • context and ontologies
  • computer-mediated communication theories
  • community/communication modelling and discourse analysis
  • computer-supported cooperative work
  • human-computer confluence and interaction
  • social network analysis
  • community participation incentives
  • emergent semantics in communities
  • social tagging systems
  • learning ontology from folksonomies
  • applications and analysis of large online communities

COMBEK will provide a forum in which practitioners and researchers can meet, and exchange research and implementation ideas and results. It will give practitioners and researchers an opportunity to present their work and to take part in open discussions. Relevant topics include (but are not limited to) theoretical or empirical exploration and position papers on theory and method, as well as tool demonstrations, realistic case studies, and experience reports.


Important Dates

Abstract Submission Deadline: June 15, 2009
Paper Submission Deadline: June 29, 2009
Acceptance Notification : August 10, 2009
Camera Ready Due: August 25, 2009
Registration Due: August 25, 2009
OTM Conferences: November 1–6, 2009

Submission Guidelines

Papers submitted to COMBEK must not have been accepted for publication elsewhere or be under review for another workshop or conference.

All submitted papers will be evaluated by at least three members of the program committee, on the basis of originality, significance, technical soundness, and clarity of expression. Submissions must be in English, and may be either full papers or short papers, both of which may discuss industrial experience or academic research. Full papers should not exceed 5,000 words (excluding references and appendices) and should not exceed 10 pages in the final camera-ready format (see later). Short position papers are vision statements describing challenging future directions, critiquing research problems, and placing new research questions. Such submissions may be up to five pages in length and will also be peer-reviewed.

The paper submission site will be announced later. Meanwhile, you can contact Pieter De Leenheer (coordinates below) for further information.

Organizing committee

Programme Co-chairs

Pieter De Leenheer (primary contact)
Vrije Universiteit Brussel STARLab, Pleinlaan 2, 1050 BRUSSEL, Belgium
+32 2 629 37 50, pdeleenh@vub.ac.be This e-mail address is being protected from spambots. You need JavaScript enabled to view it

Martin Hepp
E-Business and Web Science Research Group, Bundeswehr University, Germany

Amit Sheth
Kno.e.sis Center, Wright State University, USA

Programme Committee

• Hugo Liu, MIT Media Labs, USA
• Natalya Noy, Stanford Center for Biomedical Informatics Research, USA
• Munindar P. Singh, North Carolina State University, USA
• Dragan Gasevic, Athabasca University, Canada
• Juan Carlos Fernandez-Ramil, University of Mons-Hainaut, Belgium / Open University, Milton Keynes, UK
• Christopher Thomas, Kno.e.sis Center, Wright State University, USA
• Andreas Schmidt, FZI Karlsruhe, Germany
• Alicia Diaz, Universidad Nacional de La Plata – LIFIA, Argentina
• Tom Mens, University of Mons-Hainaut, Belgium
• Mark Aakhus, Rutgers University School of Communication, Information, and Library Studie, USA
• Filippo Lanubile, University of Bari, Italy
• Aldo de Moor, Community Sense, The Netherlands
• Igor Mozetic, Jozef-Stefan Institute, Slovenia
• Davide Eynard, Politecnico di Milano, Italy
• Tanguy Coenen, STARLab, Vrije Universiteit Brussel, Belgium
• Stijn Christiaens, STARLab, Vrije Universiteit Brussel, Belgium
• Katharina Siorpaes, SEBIS, STI, University of Innsbruck, Austria
• Marta Sabou, Knowledge Media Institute, UK
• Denny Vrandecic, AIFB, Germany
• Konstantinos Kotis, AI-Lab , research group at University of the Aegean, Greece
• Valentin Zacharias, FZI Karlsruhe, Germany

HERE is the full website.

The pervasive impedance mismatch between Business and IT

March 9, 2009

As been said and written: ontology evolution/engineering can learn many things from its much older brother of software evolution/engineering. We substantiate this by making a comprehensive literature study on ontology evolution, and look where techniques and principles from software evolution could fit in.

Apart from the similarities, there are also divergent assumptions that must be taken into account regarding vocabulary when modeling information as compared to modeling software. For example, the vocabulary used to label classes (PERSON), attributes (fName, lName), and methods is usually biased by the technical jargon of the software engineer resulting in typical constructs such as:

class PERSON
{
private:
	char fName[25];
	char lName[25];
	char Address[100];
	char people[MAXADDRESS];
        ...
};

Using technical vocabulary in a closed environment where such technical models are exchanged privately between colleague programmers usually poses no big semantic mismatch.

On the contrary, information (such as about persons) is to be exchanged publicly all over the organisation and beyond. The vocabulary used to model information should be understandable by business analysts, by administration staff, by business allies. Many dictionaries (Merriam-Webster, Wikipedia) agree that the a person is a kind of human being, however, for the term “person” there are various meanings, as defined by, .e.g, WordNet:

  1. S: (n) person, individual, someone, somebody, mortal, soul (a human being) “there was too much for one person to do”
  2. S: (n) person (a human body (usually including the clothing)) “a weapon was hidden on his person”
  3. S: (n) person (a grammatical category used in the classification of pronouns, possessive determiners, and verb forms according to whether they indicate the speaker, the addressee, or a third party) “stop talking about yourself in the third person”

None of these meanings are reflected in the software model above.

Concretely, information must be meaningful for any actor that can support the organisation in making informed analysis and decisions.

This reasonable but difficult claim is not proportional to how information is dealt with today. The information model is usually considered as an auxiliary part of the software model – in many cases even reduced to a dump for temporary data feeding and resulting from the code, without any consideration about the potential value of data for other purposes on the long term. Consequently, the vocabulary used to describe the information (i.e. the information metamodel) is the same as the one used by the software engineers to model software classes. In other words: it is rendered useless outside.

The semantic mismatch between the vocabulary used to describe information by technical and business actors causes a serious impediment to exchange that information meaningfully and do business properly. This is aggravated by the fact that business people have a different use of the information in mind than their IT colleagues.

In order to reduce the impedance mismatch as described above, a new equation must be found between the divergent perspectives of business and IT on the business concept types that live in the organisation, based on the following ideas:

  1. The meaning of information (read: business semantics) should be described in a vocabulary as close as possible to natural language, however still processable by machines. Fact-oriented approaches have been proven to be outstanding as natural-language inspired paradigm that can be utilised by layman domain experts to analyse and model their domain.
  2. This business semantics management process co-evolves with the software modelling process on the IT level.
  3. Different stakeholders (IT vs. business, but also e.g., sales vs. production floor) have various perspectives on the business concepts.
  4. The explicit rendering of stakeholders perspectives allows us to capture the evolution of the business semantics completely, validate the business semantics against these perspectives respectively, and finally increase overall satisfaction (see my Principles for ICT Democracy for more).

Principles of ICT Democracy: can communities themselves build their ontologies ?

March 4, 2009

The second-generation Web (2.0) is a complex socio-technical system of unforeseen growth and dynamics. On-line communities emerge and interact all around a usually self-organising manner supported by interactive applications, including bookmarking, tagging, blogging, and wikis, being developed and shared at little or no cost.  The emerging range of Semantic Web and other open technologies promises an increase in scale and maturity of knowledge sharing, achieved through collaboration and integration within and between diverse communities.

Considering the current pace of social and technological development, it seems that the transformation of the Web from a network of separately siloed applications and content repositories to a more seamless and semantically interoperable ecosystem is at hand, opening a wide range of scientific challenges and opportunities. However, while simple, the idea of the Semantic Web remains largely unrealised.

Semantic interoperability is the ability of two or more information systems or their (computerised) components to communicate data and to interpret the information in the data that has been communicated both in a meaningful manner, that is by means of an ontology that is shared by the involved information systems.

The ontologies that will furnish the Semantic Web are lacking, and those few that have been published are usually not based on consensus, and hence unreliable and not reusable beyond individual purposes. Of those domain vocabularies that are published on the Web, only some of them are actively maintained and thus reflect the current domain. Many others are rather outdated prototypes, not “usable and reusable”, and unworthily categorised ontologies as an agreement on the schema vocabulary is non-existing. Current techniques that claim to create semantic interoperability are unsatisfactory, both theoretically and as far as the quality of the results is concerned. In current ontology engineering practice, the underlying methodological pinciples are mostly ignored.

In current ontology engineering practices, the underlying methodological principles are mostly ignored. Furthermore, they systemically disregard the subtle gap that looms between knowledge sharing among people at the community/social level on the one hand; and information exchange between computer systems at the operational/technical level (see figure below).

The gap between knowledge sharing between human beings as an act of socialisation and information exchange between computer systems.

The gap between knowledge sharing between human beings as an act of socialisation and information exchange between computer systems.

Architecting community-driven internet systems will require a paradigm shift that goes beyond mere technological fits. In order to bridge the gap between the social and technical part of the community, one must put into practice the necessary activities to identify common needs from socialisation activities and bring the stakeholders together to find and ontological agreement to support these needs.

Community-based ontology evolution establishes the co-evolution of (A)  social interactions enabled by the community’s design; (B) the information systems that support them; and (C) the semantic patterns to fulfil semantic interoperability between these systems.

In order to enact this co-evolution we start from the following seven principles that every approach should keep in mind:

  1. ICT Democracy An ontology should be defined by its community, and not by a single developer.
  2. Autonomy Semantic interoperability requirements emerge autonomously from community evolution processes.
  3. Co-evolution Ontology evolution processes are driven by the changing semantic interoperability requirements.
  4. Perspective Rendering Ontology evolution processes must reflect the various stakeholders’ perspectives.
  5. Perspective Unification In building the common ontology, relevant parts of the various stakeholder perspectives serve as input for the unified perspective.
  6. Validation The explicit rendering of stakeholders perspectives allows us to capture the ontology evolution process completely, and validate the ontology against these perspectives respectively.
  7. Satisfaction Ultimately, co-evolving communities with their ontology will increase overall stakeholder satisfaction.

Community/business semantics management is our approach to enact community-based ontology evolution.

Business Semantics Management: a Case Study for Competency-centric HRM

January 16, 2009

In this article we introduce a novel approach and tool for fact-oriented business semantics management that is inspired by agile design methods. We demonstrate and validate it in a realistic case study that was carried out within the European Codrive project. Codrive’s vision was to contribute to more meaningful competency-centric human resource management. Key challenges are the uniform publication of unambiguous competency information and “time-to-competency” agility. To this end, we developed a shared and formal knowledge representation of competency domains. Stakeholders include educational institutes, public employment organisations, and industry partners from different European countries. The resulting Vocational Competency Ontology wanted to provide a candidate best practice for engineering a community-shared and reusable semantic pattern base that can be applied by all stakeholders to semantically reconcile their contextualised competency models.businesssemanticsmanagement De Leenheer, P., Christiaens, S., and Meersman, R. (2009) Business Semantics Management: a Case Study for Competency-centric HRM. In Journal of Computers in Industry: Special Issue about Semantic Web Computing in Industry. Elsevier, forthcoming

DOGMA-MESS: A Tool for Fact-Oriented Collaborative Ontology Evolution

November 14, 2008

ontologyeditorontology

Ontologies being shared formal specifications of a domain, are an important lever for developing meaningful internet systems. However, the problem is not in what ontologies are, but how they become operationally relevant and sustainable over longer periods of time. Fact-oriented and layered approaches such as DOGMA have been successful in facilitating domain experts in representing and understanding semantically stable ontologies, while emphasising reusability and scalability. DOGMA-MESS, extending DOGMA, is a collaborative ontology evolution methodology that supports stakeholders in iteratively interpreting and modeling their common ontologies in their own terminology and context, and feeding back these results to the owning community. In this paper we extend DOGMA Studio with a set of collaborative ontology evolution support modules.

De Leenheer, P. and Debruyne C. (2008) DOGMA-MESS: A Tool for Fact-Oriented Collaborative Ontology Evolution. In Proc. of On the Move to Meaningful Internet Systems 2008: ORM (ORM 2008) (Monterrey, Mexico), LNCS 5333, Springer, pp. 797-806

Ontology Evolution: State of the Art and Future Directions

January 19, 2008

The research area of ontology engineering seems to have reached a certain level of maturity, considering the vast amount of contemporary methods and tools for formalising and applying knowledge representation models. However, there is still little understanding of, and support for, the evolutionary aspects of ontologies. This is particularly crucial in distributed and collaborative settings such as the Semantic Web, where ontologies naturally co-evolve with their communities of use. For managing the evolution of single ontologies, established techniques from data schema evolution have been successfully adopted, and consensus on a general ontology evolution process model seems to emerge. Much less explored, however, is the problem of evolution of interorganisational ontologies. In this “complex” and dynamic setting, a collaborative change process model requires more powerful engineering, argumentation and negotiation methodologies, complemented by support for context dependency management.. It turns out that much can be learned from other domains where formal artefactsucps are being collaboratively engineered. In particular, the field of system engineering offers a wealth of techniques and tools for versioning, merging and evolving software artefacts, and many of these techniques can be reused in an ontology engineering setting. Based on this insight, this chapter gives a unified overview of the wide variety of models and mechanisms that can be used to support all of the above aspects of ontology evolution. The key remaining challenge is to construct a single framework, based on these mechanisms, which can be tailored for the needs of a particular version environment.

De Leenheer, P. and Mens, T. (2008) Ontology Evolution: State of the Art and Future Directions. In Ontology Management for the Semantic Web, Semantic Web Services, and Business Applications, from Semantic Web and Beyond: Computing for Human Experience”, eds. Hepp, M., De Leenheer, P., de Moor, A., and Sure, Y.,Springer

Business Use Case: Ontologising Competencies in an Interorganisational Setting

January 16, 2008

In this book chapter, we summarise findings from Codrive, a large-scale ontology project in the vocational training domain. This specific competency area is complex, and in order to achieve proper interoperability, all involved stakeholders must participate in interorganisational ontology engineering. In particular, this chapter illustrates the DOGMA-MESS methodology, a community-driven approach to ontology management. It presents practical experiences for the issues addressed in the previous chapters, complementingthem with illustrative data and hands-on knowledge.

Christiaens, S., De Leenheer, P., de Moor, A., Meersman, R. (2008) Business Use Case: Ontologising Competencies in an Interorganisational Setting. In Ontology Management: Semantic Web, Semantic Web Services, and Business Applications, from Semantic Web and Beyond: Computing for Human Experience, eds. Hepp, M.; De Leenheer, P.; de Moor, A.; Sure, Y., Springer