The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture




Скачать 113.55 Kb.
НазваниеThe Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture
Дата21.10.2012
Размер113.55 Kb.
ТипДокументы
The Use of UML as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture

François Pinet1, Catherine Roussey2, Thomas Brun1, Frédéric Vigier1

1 Cemagref, Clermont Ferrand, France

{francois.pinet, thomas.brun, frederic.vigier}@cemagref.fr

2 LIRIS CNRS, University of Lyon, France

catherine.roussey@liris.cnrs.fr

Abstract. For the past 20 years, ontologies have become more and more popular in various research fields such as web technologies, databases, information retrieval methods, etc. The first goal of this paper is to answer general questions about ontologies, such as: What exactly is an ontology? What is the purpose of ontology? Which types of systems use an ontology? The second goal of the paper is to help readers understand how UML can be used to model ontologies in agricultural systems. UML and the Web Ontology Language (OWL) are compared, and an example inspired from the French project named Farm Information Management is presented.

1. What is an ontology?


This section will briefly describe the history of the word “ontology” starting from Aristotle's metaphysics to the more sophisticated web technologies. Thus, readers will be able to gain a rapid overview of the use of ontologies in different domain. Then we will conclude by giving a general definition of ontologies.

Aristotle's definition of the word Ontology (with a capital O) is “the science of being qua being” (Aristotle's Metaphysics). This science is part of Philosophy, dealing with the descriptions of existing entities. Its goal is to define the general categories or primitives used to classify all the entities in the world such as human beings, animals, plants, etc. In the early 80’s, Artificial-Intelligence researchers borrowed the term “Ontology” from the field of philosophy. Ontologies became the definition of domain knowledge. They provide the possibility to separate domain knowledge from operational knowledge [29].

An interesting definition of ontologies in the field of Artificial-Intelligence was proposed by Gruber in 1993: an ontology is “the specification of conceptualisations used to help programs and humans share knowledge” [13].

He developed a more precise definition: "…ontology is a formal explicit specification of a shared conceptualization". According to Gruber, conceptualization refers to an abstract model of phenomena in the world after identifying the relevant concepts of these phenomena. Explicit means that the type of concepts used, and the constraints on their use are explicitly defined. Formal refers to the fact that the ontology should be machine readable. Shared reflects the idea that ontology should capture consensual knowledge accepted by the communities [14].

In database and information system areas, ontologies are used to facilitate the interoperability of heterogeneous information sources. The ontology is the general schema organising all the entity properties described in a group of database schemas or information sources. Each information source is provided with a wrapper that maps its database schema to the ontology. Thus, users can query the ontology and obtain a result integrating all information sources needed to answer the query [34].

Information retrieval techniques use linguistic ontology such as indexing vocabulary in order to avoid word semantic ambiguity. Thus, document and query contents are represented by concepts (i.e. the meaning of terms) and not words (i.e. sets of characters). This technique means it is possible to improve the description of document (and query) contents, and also system performance. One of the first operational systems was Ontoseek, which uses Wordnet taxonomy to describe yellow pages [20][15].

The Semantic Web is an adaptation of previous technologies. Tim Berners-Lee, the inventor of the web, defines the Semantic Web as “a web of data that can be processed directly and indirectly by machines.” Indeed, the Semantic Web is about several things: the exchange and processing of data, documents and services. The Semantic Web will integrate a community of agents capable of exchanging data and services from diverse sources in order to achieve a specific goal. More precisely, such web services are reusable software components that implement a discrete functionality (like hotel reservation) accessible through the web. Moreover document description and content will be defined precisely thanks to series of data called metadata stored in ontologies. Thus, web document retrieval is improved by this technique. The W3C1 is in charge of developing a set of standards necessary for semantic web technology [30]. All these standards are based on the XML language (W3C consortium semantic web activity).

In the agricultural domain, the well-known AGROVOC thesaurus is used to develop the Agricultural Ontology Service (AOS) project [1]. AGROVOC is a multilingual thesaurus concerning forestry, fisheries, food, environment and related domains. As presented in [1], "it consists of words or expressions (terms), in different languages and organized in relationships (e.g. 'broader', 'narrower', and 'related'), used to identify or search resources". AGROVOC was developed by the FAO and the Commission of the European Communities, in the early 1980s.

Using the knowledge contained in AGROVOC, "the AOS will be able to develop specialized domain-specific terminologies and concepts that will better support information management in the Web environment. A key objective is to add more semantics to the thesaurus, for example, by expanding and better specifying the relationships between concepts." (source: [1]). For example, the term "pollution" is associated with the term "pollutants" by the relationship "Related Terms" in the AGROVOC thesaurus. This means that the term "pollution" is related to the term "pollutants". It is possible to develop a more specific terminology. For instance, we can explicitly indicate that "pollution" is caused by "pollutants" in using a more meaningful relationship such as "caused by" [35].

To conclude, an ontology can take on various forms, and the use of the ontologies is also different, but we can state that an ontology should primarily contain:

  • a vocabulary of terms,

  • a set of term definitions which identify concepts and fix the term interpretation,

  • a modelling of the domain of interest in order to represent relationships between concepts,

  • an agreement of a community of ontology users about term definitions and the domain structure.

2. UML as an ontology language

The most common formalisms used to represent ontologies in the Artificial-Intelligence community are the Knowledge Interchange Format (KIF) [22] and Description Logics languages [9][21]. KIF is a computer-oriented language for the interoperability of knowledge among heterogeneous programs. It is based on the first-order logic and allows the formal definition of objects, functions, and relations. Description logics (DL) are a family of knowledge representation languages. They are formal and can be used to model the knowledge of an application domain. The name "logic" refers to the logic-based semantics given to these languages. Until 1980s, Description Logics were called terminological systems or concept languages. In the field of Semantic Web, the W3C proposes a language to define Ontology called OWL (Web Ontology Language) [31] developed as a follow-on from RDF (Resource Description Framework [32][4]). OWL is composed of three increasingly-expressive sublanguages: OWL Lite, OWL DL, and OWL Full. OWL Lite supports user's basic needs. The part of the OWL language called OWL DL can be viewed as an XML representation of Description Logics languages. OWL Full provides a very complete language; it is meant for users who want maximum expressiveness.

These types of languages provide interesting capabilities to develop a consistent set of concept classes; they support efficient reasoning and inference mechanisms to deduce new knowledge and compute the logical correctness of the conceptual structure. Unfortunately, these languages are little known outside Artificial-Intelligence Laboratories. Note that reasoning software are not able to support complete reasoning for every feature of OWL Full [31].

The Unified Modeling Language (UML) is a standard for modelling computational systems. It is used successfully in information system and object-oriented developments. As presented in [8][19][27][18][11], some researchers propose to use UML as Ontology language. These studies have acknowledged the benefits of using a standard modelling tool such as UML in ontology construction.

The use of the Unified Modelling Language (UML) for ontology construction has only recently been present in specialized literature. There are several common features between UML and Ontology-based languages. One main advantage of UML is that is widely used. Several tools are available for UML so designers can use them for describing their diagrams. UML is an open standard, and it is taught in many Universities. Users and developers are likely to be familiar with UML notations than KIF, Description Logics or OWL. Moreover, while KIF, DL and OWL have a textual representation, UML proposes several visual representations: class diagram, objects diagram, etc.

There are several common features between UML and Ontology-based languages [18] but the main drawback to UML is its lack of formal semantics; the traditional languages used to describe ontologies have formal semantics. The official document defining the semantics of UML contains an informal description in English [23]. Some researchers propose a mathematical model for UML (see for instance [5]). The work of Guizzardi [16][17] concerns the development of different methodological tools (UML profiles, design patterns) in order to build an ontology using UML formally and correctly. Cranefiel and Purvis develop an ontology using UML and Object Constraint Language (OCL) [24]; the concepts are described by UML classes and constraints of concepts are described in OCL [8].

Because UML is widely used and supported by numerous tools, it could be viewed as a good candidate to develop ontologies. Nevertheless, UML is a semi-formal formalism and will not be sufficient to represent all the details required by complex reasoning processes [7]. OCL is sometimes considered as “a variant of [first order predicate logic] tuned for writing constraints on object structures” [6], and might overcome this limitation in the future; in the OMG specification of OCL [24], an annex presents a first version of a formal semantic. Currently, this annex doesn't describe how to deduce new knowledge, or compute the logical correctness of OCL constraints.


3. Similarity and differences between UML and traditional languages used to describe ontologies

Several studies compare UML and formal languages used to describe ontologies [7][18][3]. OWL is usually considered as the flagship language for ontologies [3][10]; thus this chapter focuses on a comparison between OWL (Full) and UML.

According to existing studies, there is a common part between UML and OWL (Figure 1). The first subsection presents the main common concepts between the two languages and the second subsection evocates several differences.

3.1 Mappings between UML and ontology languages

In the OMG recommendation [18], mappings among the metamodels of OWL and UML are proposed. Several common features between UML and OWL are shown. Table 1 presents the main similarities between UML class diagrams and OWL.




Figure 1. UML and OWL.


UML

OWL

class

class

subclass

subclass

instance

individual

attribute, association

property

mutliplicity

minCardinality,

maxCardinality,

cardinality


Table 1. OWL and UML: the common features - see more details in [18]



  • Class and subclass


UML and OWL are based on classes. In OWL, classes correspond to concepts defined in the ontology. Subclasses can be also modelled in both languages but OWL integrates a universal class Thing that generalizes all classes of a given ontology.

The class Animal of the Figure 2 is declared in OWL as follows:



The syntax rdf:ID="Animal" is used to declare a new name in the ontology (i.e. a new term). The Animal class can now be referred to as "#Animal" in the rest of the ontology.

The OWL taxonomic notation for classes is subClassOf. OWL subClassOf can be considered as similar to UML generalization relationships [18]. According to the W3C OWL Guide [31]:

"It [i.e. subClassOf] relates a more specific class to a more general class. If X is a subclass of Y, then every instance of X is also an instance of Y. (…) If X is a subclass of Y and Y a subclass of Z then X is a subclass of Z."

"A sheep is an animal" modeled in OWL:











Figure 2. An UML class diagram; Ranches have flocks of sheep, and a sheep is an animal having four legs.


  • Object / Individual


In the object-oriented paradigm, instances of a class are referred to as objects. In OWL, an instance of class is called an individual. An individual is introduced by declaring it to be a member of a class. The next example declares the sheep named "Sheep_1" (i.e. the individual "Sheep_1" of the class Sheep):









The first line declares an individual "Sheep_1"; the other lines define that this individual is a member of the class Sheep.


  • Attribute, Association / Property


Class attributes (in UML) can be modeled with properties in OWL. Properties relate individuals to individuals (object properties), or individuals to data types e.g. integer, string, etc. (datatype properties).

The next example illustrates datatype properties. The class Sheep_Ranch and its attributes (presented in Figure 2) are declared in OWL as follows:



















The datatype properties are defined by a domain and a range. If we compare with UML, the datatype properties are attributes, the domains are their classes, and the ranges are their types. In the example, the presented ranges ("&xsd:integer" and "&xsd:string") belong to datatypes recommended by W3C. They correspond to XML Schema datatypes [33].

Links between individuals can be created in using object properties. For instance, it is possible to add an object property "hasFlock" to the class Sheep_Ranch.









The range of this property is the class Sheep, i.e. the values of the property hasFlock are instances of the class Sheep. Thus, a link is created between the instances of the class Sheep_Ranch and the instances of the class Sheep. In UML, this object property can correspond to:

  • an attribute hasFlock in the class Sheep_Ranch; the type of the attribute is "Sheep", or

  • an association end hasFlock between Sheep_Ranch and Sheep (as shown in Figure 2).

The inverse of a property is modelled with inverseOf e.g. the inverse of hasFlock is modelled as follows:











The inverse of hasFlock is hasRanch. The range of hasRanch is the class Sheep_Ranch i.e. the values of the property hasRanch are instances of the class Sheep_Ranch; inverseOf is used to declare that hasRanch is the inverse of hasFlock. Thus, the both association ends ("hasFlock hasRanch") of the UML diagram are modeled in OWL.


  • Multiplicity / Cardinality


In UML, multiplicities are important to constrain the number of objects linked by associations. In OWL, any instance of a class may have an arbitrary number (zero or more) of values for a particular property [31]. To allow only a specific number of values for that property, cardinality constraints can be used. OWL provides the concepts cardinality, minCardinality, and maxCardinality for restricting the cardinality of properties. For example, we model that a sheep ranch has at least one sheep:





1





The cardinality constraint minCardinality belongs to the value space of the XML Schema datatype nonNegativeInteger. A minCardinality constraint describes a class of all individuals that have at least N semantically distinct values for the property concerned.

Let Leg be a class, and hasLimb be a property of the class Sheep i.e. a domain of hasLimb is Sheep. The following example constrains the number of values of the property hasLimb:





4





This cardinality restriction is similar to the multiplicity of the association hasLimb in the UML diagram shown in Figure 2.

3.2 Differences between UML and ontology languages

The author of [3] highlights the general differences between an ontology and an UML model. The main results of this study are presented in Table 2.



Ontology

UML Model

- originated from the Artificial

Intelligence world for the

purpose of precisely capturing

“knowledge”

- originated from the software

engineering world for the

purpose of simplifying the

modelling of software


- OWL is the main language


- UML/OCL are the languages

used to specify the models


- formal semantics

(Description Logics)


- semi-formal semantics

(UML semantics are expressed

by a metamodel)



Table 2. General differences between an ontology and an UML model [3].



  • In OWL but not in UML


OWL allows defining a class as the set of individuals which satisfy "restrictions" e.g. a sheep is an animal belonging to the flock of a sheep ranch and having four legs. It is possible to infer from the properties of an individual that it is a member of a given class. OWL provides reasoning and inference mechanisms to deduce knowledge (e.g. determine that an individual belongs to a specific class) and compute the logical correctness of the conceptual structure (e.g. determine that the ontology is consistent).

Thus, it becomes possible to deduce that an animal is a sheep because it satisfies several properties: for example, it has four legs and it is belongs to the flock of a sheep ranch. Currently, OCL can be used to model constraints (i.e. restrictions) on UML diagrams, but it doesn't integrate formal inference mechanisms.



  • In UML but not in OWL


UML allows the specification of behavioural features. UML is able to model operations, parameters of methods, interface classes, etc. The interaction diagrams, activity diagrams and state diagrams can be also used to describe precisely dynamic aspects: sent messages, different states and life lines of objects, etc.

An ontology is mainly a model presenting relationships between concepts. Thus, OWL and the other ontology languages do not provide support for expressing the behaviour of a system. This is the reason why only UML class diagrams and object diagrams are used to model ontologies.


Because UML class and object diagrams are widely used, it could be viewed as a good candidate to model visually the main features of ontologies i.e. classes, subclasses, properties, individuals, etc. (see Figure 1 and Table 1). As presented in the next section, in the context of a French project, UML class diagrams have been used to specify ontologies in order to model consensual knowledge of the agricultural domain. The visual representations provided by UML had facilitated the design process of the ontology. The goal of the project was to define a standard to build a data interchange format for French agricultural information systems [12].


4. Farm Information Management Project

4.1 Exchange of agricultural data: a need which is partially met

In France, the main economic and institutional players have made commitments to developing diverse systems and standards, making it possible to control and manage the increasing flow of information linked to farm activities. Their main goal is to facilitate the interoperability between information systems.

As an example, the French Data Reference Centre for Water (SANDRE in French)2 is in charge of developing a common language for water data exchange [26]. Data related to water in France are issued from thousands of organisations and public services. The SANDRE's priorities are to make compatible and homogeneous data definitions between producers, users and databanks. For example, some themes considered by SANDRE are: groundwater, hygrometry, fertiliser spreading, etc. SANDRE proposed "a common language concerning data involved in the French Water Information System. Specific terms relevant to water data are clearly defined and data exchange specifications are also produced to fulfil the communication needs between partners involved in the field of water" [26]. One of the SANDRE’s goals is to define, at a national level, a common vocabulary concerning the field of water (SANDRE’s common language). To fulfil this task, data models have been developed. They are associated to data dictionaries that gather all the definitions of data related to topic Water. XML-based exchange formats have been also proposed. The application range of SANDRE is larger than the farm activities.

In another context, the association AgroEDI3 Europe led several working groups to propose standards in agriculture for specific productions or activities (e.g. viticulture, phytosanitary treatments) [2].

These actions contribute to the development of professional information exchange and open the way to electronic management; they contribute to the progressive advent of a farming “network” by proposing several data interchange formats in some cases facilitating communication between certain types of agricultural information systems. However, the various actions that have been carried out have led to the definition of standards for each production chain or for each field of activities (e.g. water data management). A more general approach was necessary in order to propose a unified standard that covers all the main concepts used in the majority of agricultural information systems.

4.2 Ontology definition in agriculture: a means of communication

Thus the members of the new project named FIM4 have studied a unified standardisation approach. The standard is not only limited to a production chain or a type of activities as in previous approaches; it can be used in the majority of contexts in agriculture. The standard covers the common concepts used in the main production chains and is designed to be used in different fields of activities in agriculture in France. A final goal of the standard is to provide more complete data interchange formats in order to facilitate and to improve interoperability between agricultural information systems.

The first task of the project teams was to carry out an inventory of the various initiatives dealing with their themes. Then, different terms, concepts and their relationships have been identified for each theme. An important part of the FIM project consists in integrating and enhancing the definition of concepts, and work on standardisation already initiated by the various partners. The monitoring of these approaches and the participation in various work groups and their corresponding project committees are therefore fully integrated in the project.

An ontology has been chosen to formalize the standard. All the members of the project can propose new concepts to the developed ontology. Data interchange formats are also proposed on the basis of the vocabularies and the concepts of the ontology. The ontology is represented by UML class diagrams. UML has been chosen to model the ontology because the participants of the FIM project are familiar with UML. Figure 3 summarizes the presented approach.

The developed ontology is composed of several parts. Each of one is related to a specific theme [12]:

  • Actors (modelling the main features of the actors involved in the agricultural production - farms, farmers, etc.),

  • Foods (for animals),

  • Analysis and monitoring of the production,

  • Soil analysis,

  • Buildings,

  • Animals,

  • Works related to the agricultural production (agricultural spreading, harvest, etc.),

  • Identification and traceability of animals,

  • Agricultural inputs (fertilizer, water, etc.)

  • Stockbreeding,

  • Regulations,

  • Land use,

  • Crops,

  • Livestock reproduction,

  • Animals and sanitary monitoring.





Figure 3. From standard to Data Interchange Format


4.3 Use of UML

This section shows some examples taken from the FIM project. They illustrate certain concepts of the ontology represented with UML.

Figure 4 is based on the FIM project theme “Animals and Sanitary Monitoring”. This theme deals with all concepts related to animal health: every curative and preventive measure taken for animals. This example presents the main concepts for sanitary event management (Animals, Interventions, Medical Treatments, Sanitary Products, Actors, etc.).

The class Sanitary Event is the main concept of the diagram. It may correspond to surgical interventions, administrations of drugs, dietary treatments. Sanitary events concern animals and are associated to different categories of products: biocide, veterinary products, etc. Actors can be involved in the events (veterinaries, technicians, institutions). A medical treatment can be viewed as an aggregation of sanitary events.


Figure 4. Animals and sanitary monitoring


Figure 4. Animals and sanitary monitoring


Only the basic UML notations have been used (class, attribute, relationship). Each concept is represented by a class. Only the main attributes are on the diagram. Hierarchies are used to model the different types of interventions, sanitary products, and actors. These nomenclatures are represented in UML using the generalisation/specialisation relationship. This gives more flexibility to the model, providing the possibility of adding a new component for instance, and/or making existing concepts more specialized.




Figure 5. Land use


The second example is an extract from the theme “Land Use” (Figure 5). It illustrates the use of hierarchies. It models that a farm is composed of agricultural parcels; the land use of their parcels is represented by a hierarchy (Cultural Land Use, Infrastructure, Building).

A documentation associated to classes and attributes is also described in natural language [12]. Thus, the ontology modelled in the FIM project presents:

  • a vocabulary of terms,

  • a set of term definitions which identify the concepts and fix the term interpretation,

  • the relationships between concepts.

5. Conclusion and Perspectives


In the context of the FIM project, UML is used to formalize the different concepts common in agriculture (see [12]). The proposed ontology captures consensual knowledge accepted by the experts. UML enables the participants of the FIM project to discuss the concepts easily. Subsequently, these concepts are used to propose a standard data interchange format in order to facilitate and to improve the interoperability between agricultural information systems in France.

An interesting perspective will be to translate the ontology described in UML into a formal ontology (RDF, OWL, etc.) in order to offer new possibilities: produce reasoning, reach the requirement of the semantic Web, integrate several database schemas, etc. In this case, one solution consists in translating the UML diagrams into a formal ontology language, and then enriching the produced formal ontology. For example, it is possible to convert UML specifications into a formal ontology with Protégé and its UML Storage Backend Plug-In [28]. Then, additional specifications can be integrated into the ontology: creation of individuals, modelling of inferences, etc. For instance, the works of [25] propose to start modelling an ontology with a UML class diagram. After that, this UML specification is translated into OWL with Protégé and its UML Storage Backend Plug-In. Then, inferences are defined with Description Logic in order to deduce new knowledge. A DIG reasoner is used to produce new information. A DIG reasoner supports the standard interface defined by the Description Logic Implementation Group (DIG - http://dl.kr.org/dig/). This standard provides a specification for connecting DL reasoners and different ontology modelling tools such as Protégé. A DIG reasoner provides an access interface which enables the reasoner to be accessed over HTTP.


References

1. AGROVOC – A multilingual Agricultural Thesaurus - ftp://ftp.fao.org/gi/gil/gilws/aims/references/flyers/agrovoc_en.pdf

2. Agro-EDI Europe: Agro-EDI Assocation Web Site: http://www.agro-edi.asso.fr

3. Atkinson C.: Models versus Ontologies - What's the Difference and where does it Matter? VORTE2006, Regal Kowloon Hotel, Hong Kong, October 16th 2006

4. Bekett, D.: RDF/XML Syntax Specification 2004 - http://www.w3c.org/RDF

5. Breu, R., Grosu, R., Huber, F., Rumpe, B., Schwerin, W.: Towards a precise semantics for object-oriented modeling techniques. In Object-Oriented Technology, ECOOP'97 Workshop Reader, 1997

6. Clark T., Evans A., Kent S., Brodsky S., Cook S.: A feasibility study in rearchitecting UML as a family of languages using a precise OO meta-modeling approach. http://www.cs.york.ac.uk/puml/mmf/, 2000

7. Cranefield, S.: UML and the Semantic Web, In: the International Semantic Web Working Symposium, Palo Alto, 2001

8. Cranefield S., Purvis., M.: UML as an ontology modelling language. In Proceedings of the Workshop on Intelligent Information Integration, 16th International Joint Conference on Artificial Intelligence (IJCAI-99), 1999

9. Donini, F., Lenzerini, M., Nardi, D., Schaerf, A.: Reasoning in description logics. In Gerhard Brewka, editor, Principles of Knowledge Representation, Studies in Logic, Language and Information, pages 193-238. CSLI Publications, 1996

10. Drummond N., Shearer R.: The Open World Assumption or Sometimes its nice to know what we don’t Know, University of Manchester - http://www.cs.man.ac.uk/~drummond/presentations/OWA.pdf

11. Gasevic D., Djuric D., Devedzic V.: Model Driven Architecture and Ontology Development, Springer, 328p, 2006

12. GIEA: GIEA (FIM) Web site: http://www.projetgiea.fr

13. Gruber, R.: Towards Principles for the Design of Ontologies Used for Knowledge Sharing. In International Workshop on Formal Ontology, Padova, Italy, 1993. Available as technical report KSL-93-04

14. Gruber, R.: A Translation Approach to Portable Ontology Specifications. Knowledge Acquisition vol. 5(2), pp. 199-220, 1993

15. Guarino, N., Masolo, C., Vetere, G.: OntoSeek: Content-Based Access to the Web, IEEE Intelligent Systems vol. 14(3), 70-80, 1999

16. Guizzardi, G.; Falbo, R.A.; Pereira Filho, J.G.: Using objects and Patterns to implement domain ontologies, Journal of Brazilian Computer Society (JBCS), Special Issue on Software Engineering vol. 8(1), 2002

17. Guizzardi, G.; Wagner, G.; Guarino, N.; van Sinderen, M. “An Ontologically Well-Founded Profile for UML Conceptual Models”, Lecture Notes in Computer Science vol. 3084, 2004

18. IBM: Ontology Definition Metamodel, June 2006

19. Philippe, M.: Translations between UML, OWL, KIF and the WebKB-2 Languages (For-Taxonomy, Frame-CG, Formalized English), Technical Report, May/June 2003

20. Miller, G., Beckwith, R., Fellbaum, C., Gross D., Miller, K.: Introduction to WordNet: an on-line lexical database'. International Journal of Lexicography vol. 3 (4), 1990, Revised August 1993 - ftp://ftp.cogsci.princeton.edu/pub/wordnet/5papers.ps

21. Nardi, D., Brachman, R. J.: An Introduction to Description Logics. In the Description Logic Handbook, edited by F. Baader, D. Calvanese, D.L. McGuinness, D. Nardi, P.F. Patel-Schneider, Cambridge University Press, 2002

22. National Committee for Information Technology Standards, Technical Committee T2 (Information Interchange and Interpretation): Knowledge Interchange Format,draft proposed American National Standard (dpANS). NCITS.T2/98-004 -http://logic.stanford.edu/kif/dpans.html

23. Object Management Group: Unified Modeling Language, version 1.5, Mars 2003 - http://www.omg.org/docs/formal/03-03-01.pdf

24. Object Management Group: UML 2.0, OCL specification, May 2006

25. Pinet, F., Ventadour, P., Brun, T., Papajorgji, P., Roussey, C., Vigier, F.: Using UML for Ontology Construction: a Case Study in Agriculture. In: the 7th AOS Workshop on Ontology-Based Knowledge Discovery: Using Metadata & Ontologies for Improving Access to Agricultural Information, Bangalore, India, November 2006

26. SANDRE: SANDRE Web site: http://sandre.eaufrance.fr/rubrique.php3?id_rubrique=60&lang=en

27. Schreiber, G.: A UML Presentation Syntax for OWL Lite, Technical Report, 2005

28. Standford University: Protégé, 2005 - http://protege.stanford.edu

29. Studer, R., Benjamins, V., and Fensel, D.: Knowledge engineering: Principles and methods. IEEE Transactions on Data and Knowledge Engineering vol. 25: p. 161-197, 1998

30. W3C: Semantic Web, 2001 - http://www.w3.org/2001/sw/

31. W3C: OWL Web Ontology Language:Overview, W3C Recommendation 2004-2-10 - http://www.w3.org/TR/owl-features/

32. W3C: Resource Description Framework (RDF): Concepts and Abstract Syntax W3C, W3C Recommendation February 2004, http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/

33. XML Schema Part 2: Datatypes Second Edition, W3C Recommendation, October 2004, http://www.w3.org/TR/xmlschema-2/

34. Wiederhold G.: Mediators in the Architecture of Future Information Systems. IEEE Computer 25(3): 38-49 (1992) 8

35. Wikipedia: AGROVOC - Web Site: http://en.wikipedia.org/wiki/AGROVOC


1 World Wide Web Consortium – www.w3c.org

2 The SANDRE is composed of a member from one of each signatory organisations of the "French Information System for Water" [26]: the French Ministry of Ecology, French Institute for Environment, Water Agencies, Fishing Council, French Research Institute for Exploitation of the Sea, "Electricité de France", Institute for Geological and Mining Researches, International Water Institute.

3 EDI : Electronic Data Interchange

4 Farm Information Management (GIEA in French). The list of the different participants involved in the FIM project can be found at: http://www.projetgiea.fr/-Organisation-

Похожие:

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconArchitecture Independent Parallel Design Tool: a refinement-Based Methodology for the Design and Development of Parallel Software

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconUnit1: Design of Machine Tool Drives

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconAn Object-Oriented Support Tool for the Design of Casting Procedures

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconStresses, Stress tensor, Principal Stresses. Engineering Materials and their mechanical properties, Stress-Strain diagrams, Stress Analysis, Design considerations: Codes and Standards. 05 Hours unit- 2 design for static & impact strength

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconSummary of evidence on proposed housing design standards for the

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconApplying uml and Patterns: An Introduction to Object-Oriented Analysis and Design and Iterative Development (3rd Edition)

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconTool Use: Misuse of shop tools will result in the loss of tool privileges

The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconЛабораторная работа №7
Изучение приложения siso design Tool пакета Control system системы matlab для моделирования, имитирования и анализа систем управления...
The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconВопросы к экзамену по управлению предприятием
Объектно-ориентированные методы проектирования. Нотация uml. Виды моделей нотации uml
The Use of uml as a Tool for the Formalisation of Standards and the Design of Ontologies in Agriculture iconKeywords: graphic design history, envisioning, information design, timeline design

Разместите кнопку на своём сайте:
Библиотека


База данных защищена авторским правом ©lib.znate.ru 2014
обратиться к администрации
Библиотека
Главная страница