=Paper=
{{Paper
|id=None
|storemode=property
|title=Semantic UBL-like documents for innovation
|pdfUrl=https://ceur-ws.org/Vol-1006/paper7.pdf
|volume=Vol-1006
|dblpUrl=https://dblp.org/rec/conf/caise/BasarSST13
}}
==Semantic UBL-like documents for innovation==
https://ceur-ws.org/Vol-1006/paper7.pdf
Semantic UBL-like documents for innovation
R. Volkan Basar1, A. Anil Sinaci1, Fabrizio Smith2, Francesco Taglino2
1
SRDC Software Research & Development and Consultancy Ltd.
Silikon Blok No:14, Teknokent ODTU, 06800 Ankara, Turkey
{anil,volkan}@srdc.com.tr
2
National Research Council, IASI “Antonio Ruberti”
Viale Manzoni 30, 00185 Roma, Italy
{smith,taglino}@iasi.cnr.it
Abstract. Considering innovation as the result of only spontaneous activities is
a simplistic vision, because working out inspiration and reaching up to
innovation need awareness and knowledge about the application domain and its
problems. In this paper, we address the issue of knowledge representation,
access and sharing in an enterprise context, by proposing an ontology-based
framework (DocOnto) for the semantic description of documents involved in
innovation activities. The framework, which is built within the BIVEE
European project, is characterized by a customizable approach inspired by the
UBL/CCTS, which allows each enterprise to refine the DocOnto at best for its
needs. Then UBL-like structures are semantically lifted and used for describing
concrete documents. Such a semantic representation enables reasoning services
like querying and retrieving of documents, understanding similarities among
documents, assessing their status and quality, monitoring innovation activities.
The framework is supported by the technological integration of the iSurf
eDoCreator, for modelling UBL-like documents structures, and the Production
and Innovation Knowledge Repository (PIKR), the semantic knowledge hub of
the BIVEE platform.
Keywords: business innovation, ontologies, semantic description, UBL, CCTS,
document management.
1 Introduction
“Genius is one percent inspiration and ninety-nine percent perspiration”. This
quotation from Thomas. A. Edison reveals what is behind innovation successes.
Innovation is often identified as the result of spontaneous attitudes like creativity or
artistic flair, but bringing new ideas into the market in the form of innovative products
or services, also needs the adoption of methods and supporting means. In particular,
in the era of the information society, knowledge management has a primary role also
in relation to innovation activities. With respect to that, [1,4] consider innovation as a
practice and process that captures, acquires, manages and diffuses knowledge with the
aim to create new knowledge. Furthermore, knowledge enables creativity by
permitting knowledge associations and linkages that otherwise are difficult to be
discovered.
In this paper, we outline a framework for designing semantics-based structures
(Document Ontology, DocOnto) to enable the semantic enrichment and management
of innovation-related documents, i.e. documental resources produced and consumed
during innovation initiatives (e.g., proposed ideas, feasibility studies, etc.). This
activity is conducted within the BIVEE1 European project, which is about the
development of an ICT infrastructure for supporting innovation activities in virtual
enterprise (VE) environments.
Among related initiatives, we mention Dublin Core2, a vocabulary of fifteen
properties for description of documental resources, and SALT [8], which is for
describing the organization of a document in terms of sections and paragraphs. While
we intend to re-use part of the terms from Dublin Core, we look at documents
differently from SALT, since we are focused on the semantics instead of the
organization of the structure of a document.
The proposed framework is based on the one hand, on the methodological
integration of the UBL/CCTS [7] approach, which is for modeling and customizing
documents structures, together with semantic representation methods. On the other
hand, the framework is supported by the technological integration between the iSurf
eDoCreator[6] UBL editor and the Production and Innovation Knowledge Repository
(PIKR), which is the semantic knowledge hub of the BIVEE platform. The objective
is the semantic lifting of innovation-related documents structures and content for
enabling interoperability and openness, as well as reasoning services such as querying
and retrieval of documents, reasoning over documents description, understanding
similarities among documents, assessing status and quality of documents and,
monitoring innovation activities.
The paper is organized as follows. Section 2 presents the overall structure of the
DocOnto and the UBL-inspired approach for building and customizing the DocOnto
itself. Section 3 focuses on the technical aspects concerning the integration between
the PIKR and eDoCreator, as well as on the services provided by the PIKR for
reasoning over the semantic description of annotated documents. Conclusions and
future work end the paper.
2 The DocOnto framework
One of the main objectives of the BIVEE project is to support and facilitate
innovation activities in a VE environment. To this end, the Virtual Enterprise
Modeling Framework (VEMF) has been developed. According to the VEMF,
innovation-related activities happen within four waves: Creativity, Feasibility,
Prototyping and Engineering. Flowing through these four waves, many documents are
produced, used, consumed and evaluated. For instance, in the Creativity wave, given a
problem or issue, many ideas can be proposed to address it. Some of them will pass
1
Business Innovation and Virtual Enterprise Environment (No. FoF-ICT-2011.7.3-285746).
2
http://dublincore.org/documents/dces/.
the initial stage and will be further elaborated. Recording such information means
keeping track of reasons that guide decisions, and re-using knowledge to save time
and money in the future. We consider that ontology-based semantic technique can be
effective in addressing representation, sharing, access, and reasoning over documental
resources, especially in VE context where boundaries are larger and a reference
(ontology) is requested.
For the definition of the proposed ontology-based innovation document framework
(DocOnto) we started from the results of an activity performed within the BIVEE
project: two end-user organizations were asked to see their innovation-related
activities through the four innovation waves and indicate the information they actually
produce and use. This brought the identification of sets of documents, one for each
end-user [5]. These results have been taken as specifications and, starting from them,
a conceptualization of these documents has been performed for identifying valuable
InfoItems (building blocks, which correspond to small and meaningful elements),
InfoSets (recursive aggregation and association of InfoItems) and associations
between them as described below and reported in Table 1.
Header groups meta-data InfoItems like the title of the document, the authors etc.
Content groups InfoItems describing the essence of the document, i.e., its
semantics. The adoption of domain-focused dictionaries, thesauri or ontologies
increments the level of interoperability and enables reasoning mechanisms.
Related Knowledge Resources section allows to establish relations between
InfoSets such as 'prerequisite', 'feedback', 'partOf', 'relatedTo' etc.
Header
Title Advanced HMI
Description System for the robot programming based on the 3d reconstruction of
the inspected components
... ...
Content
Research Line 3D vision, cloud point, artificial intelligence algorithm
Technology HMI
... ...
Related Knowledge Resources
Part of doc:IP_AdvancedHMI
Table 1. An example of InfoSet instance: a Technical Solution Report
2.1 UBL Customization Approach
The UN/CEFACT Core Component Technical Specification (CCTS) has the notion of
building blocks, called Core Components (CC). Core components are context-neutral
having a generic semantics and purpose, and can be re-used in different contexts [6].
Business Information Entities (BIE) are contextualized CCs and have three types:
Basic Business Information Entity (BBIE), Association Business Information Entity
(ASBIE) and Aggregate Business Information Entity (ABIE). UBL [7] implements
CCTS and publishes XML based Business Document Definitions, Common BIEs and
Data Types such as an Invoice document or an Address BIE
Data requirements change for different virtual enterprises in order to address the
needs of innovation activities. Hence, it is required to customize the DocOnto for each
virtual enterprise once the requirements have been set. UBL provides a
methodological way for the customization of already available documents and BIEs.
Since this methodology has already been implemented by eDoCreator tool, our
solution inherently supports customization of existing innovation related documents
and BIEs. According to the UBL standard, new information entities can be added to
meet the requirements of a specific business context, optional information entities can
be omitted, the meaning of information entities can be refined, new constraints can be
specified, new aggregations or documents can be combined or assembled or new
business rules can be added during a customization. If a new type of innovation
document is required, users can model its structure through customization facilities
offered by eDoCreator, which are conformance with customization guidelines of
UBL. Since we model documents through InfoItems and InfoSets, and follow the UBL
approach, our modelling directly maps to UBL terms when we leave out the
technologies of our framework. This mapping can be depicted as follows: BBIE -
InfoItem, ABIE - InfoSet and ASBIE - Associations.
3 TECHNICAL REALIZATION
In this section we give an overview of the technical aspects related to the integration
of the PIKR [2], and the eDoCreator for supporting the implementation of the
DocOnto. We also outline the semantic services in charge of exploiting the semantic
description of the documents in terms of the DocOnto.
About the integration between the eDoCreator and the PIKR, the former exports
XML Schema3 of modelled documents to a Mediator module. The Mediator performs
the semantic lifting by encoding documents structures into OWL/RDF 4, the de-facto
standard for ontology and meta-data sharing. The result of the lifting is then
transmitted to the PIKR, which maintains it in a triple store.
The knowledge representation framework discussed in the previous sections
enables the enactment of a number of reasoning facilities to support the management
of documents in innovation projects, in terms of the following services.
Search. This service provides keyword-based search functionalities. The user request
is expressed as an ontology-based feature vector describing the criteria for the
selection of the resources of interest. By applying semantic similarity techniques (the
SemSim metrics [3]) the degree of matching among the terms used to formulate the
request and the ones used to describe the available resources is computed, and a list of
ranked results, with respect to the Semsim similarity metrics, is returned. For instance,
3
http://www.w3.org/TR/xmlschema-0/
4
http://www.w3.org/TR/owl2-overview/
suppose that the user is interested in finding all the documents that have been
authored in the last two years and concerning the initial stages of the design of a piece
of furniture equipped with an electronic device. The corresponding request should be
formulated as follows:
{content:[Furniture, Electronic_Device]; type= Proposal,
creationWave=Creativity, issueYear>2010}
The engine will retrieve semantically related resources, such as Proposed Idea or
Project Proposal documents about a Contour Chair with an embedded Media Player
(which are assumed to be defined in the domain ontology as kinds of piece of
furniture and electronic device, respectively).
Query. This service enables us to retrieve pieces of knowledge which exhibit some
given properties. Queries are posed in terms of the vocabulary and semantic relations
provided by the PIKR ontologies, and the underlying reasoning engine returns a list of
answers that satisfy all the specified properties. These answers may consist of factual
knowledge (DocOnto instances), conceptual knowledge (ontological terms), or
references to concrete resources. We are currently developing a query language, based
on SELECT-WHERE paradigm along the line of the SPARQL5 standard. For
instance, to identify reusable best practices or technical solutions in a given domain,
we may want to retrieve all the protocols related to documents addressing the research
line 3D_Vision. This can be expressed as follows.
Q(?p) : protocol(?p) AND related(?p,?doc) AND
research_line(?doc,3D_Vision)
Compliance Checking. This service allows us for checking the compliance of the
factual knowledge, captured at a given time in the semantic description of the
documents, with respect to business policies and internal regulations. Compliance
requirements can be represented in the DocOnto as business rules, i.e., statements that
define or constrain the structure of the documents or the dependencies among them on
the basis of the sequencing of business operations. The compliance check verifies the
consistency among the assertions contained in the DocOnto instances and the axioms
defined in the Knowledge Resource Ontologies formalizing the business rules.
Examples of constraints are “Each Innovation Report needs to be composed by a
Project Proposal and a Market Analysis", or "A Monitoring Sheet cannot be produced
unless a Gantt Chart has been finalized before". The former rule can be formalized by
the following axiom:
if innovation_report(x) then y,z. project_proposal(y)
and market_analysis(z) and partOf(x,y) and partOf(x,z)
5
http://www.w3.org/TR/rdf-sparql-query/
4 Conclusions and Future Work
In this paper we outlined an ontology-based framework for semantic description of
innovation-related documents. We have elaborated on CCTS and UBL approaches
and identified a bunch of InfoSets corresponding to categories of information that are
produced, consumed and evaluated during innovation projects. Furthermore, we
identified relationships that can occur among InfoSets, and we started to identify
InfoItems, elementary components of the InfoSets. We intend to re-use available
vocabularies as much as possible to enable Linked Data approach in this document
management methodology.
References
1. Cohen, W.M. and Levinthal, D.A. (1990), “Absorptive capacity: A new perspective on
learning and innovation”, Administrative Science Quarterly, Vol. 35, 128-152.
2. Diamantini, C., Potena, D., Proietti, M., Smith, F., Storti, E., Taglino, F.: A semantic
framework for knowledge management in virtual innovation factories. International
Journal of Information System Modeling and Design. To appear.
3. Formica A., Missikoff M., Pourabbas E., Taglino F. (2013) Semantic search for matching
user requests with profiled enterprises. Computers in Industry, 64: 191-202.
4. Gloet, M. and Terziovski M. (2004), “Exploring the Relationship between Knowledge
Management Practices and Innovation Performances”, Journal of Manufacturing
Technology Management, Vol. 15 No. 5, pp.402-409.
5. Sinaci A., Piersantelli M., Cristalli C., Gigante F., Laleci G., Basar V. (2012), "A
Document Centric Approach for User Requirements in BIVEE", CEUR Workshop
Proceedings Vol. 864 Article 5
6. Tuncer F., Dogac A., Postaci S., Gonul S., Alpay E. (2009), "iSURFeDoCreator: e-
Business Document Design and Customization Environment"
7. OASIS UBL TC (2006), "Universal Business Language v2.0". Retrieved March 3, 2013
from http://docs.oasis-open.org/ubl/os-UBL-2.0/UBL-2.0.pdf.
8. Groza T., Handschuh S. (2009), "Salt Document Ontology (SDO). Retrieved March 3,
2013 from http://salt.semanticauthoring.org/ontologies/sdo#.