Another issue concerns the remaining software re- The second group of issues that needed to be ad- sources, such as applications, libraries etc. In this dressed originates from the fact that our ontology has case the CGO, again, defined a rich set of classes, to include classes and properties for the description of but the defined properties were limited to installed- terms and parameters to be used in the Service Level Software which notably lacks the definition of domain Agreement SLA negotiations.
Obviously, while the and range and needLib which defines a relationship above mentioned issues required modifications to the between a GridApplication and a Library. In our CGO, here we deal with classes and properties that, for case we needed, on the one hand, a property like in- obvious reasons, were never considered to be included stalledSoftware but with range and domain restricted in the CGO.
On As described above, negotiations between agents the other hand, we also needed a way to specify the in the system materialize primarily in two main software requirements of a job that could be used to scenarios: 1 contracting a job to be executed, and search for a node that the job could be run on. For the time being we assume that the implementation of these processes will continue to 3.
Therefore, the ontology used The last of the key issues discovered when ana- in the system needs to contain definitions of required lyzing the CoreGRID ontology concerned the way it messages and their possible content. Messages that dealt with the CPU specification. Recent years are are needed by both processes should include the initial characterized by an explosion of processor types.
We call for proposals, proposal, acceptance or refusal of a are dealing not only with vector processors, and multi- proposal, and information about the final result; i. In this context it narios, negotiation will involve multiple parameters; has been noted that in many cases especially for more since in real-life situations there are multiple factors specialized jobs , there may exist very detailed speci- influencing the final decision, e. Some jobs a job to be executed price is naturally important, may, for example, be optimized for the CBE, while however, it may happen that the User is willing to others may be requested to not to be executed on vec- pay more money to have the task competed earlier.
Similarly, multicriterial analysis. Hence, for each negotiation for some CPU and memory intensive applications e. Parameters concerning joining a clearly specified domain and a strongly typed range, team include, for instance, revenue, length of contract, which allows us to use them in Java code generation. During the negotiation process messages with new properties related to them: hasArchitecture, content based on the ontology have to transfer not hasVendor, hasMemory.
Furthermore, we have added only the exact parameter values but also constraints a hierarchy of properties allowing for software config- to be imposed on specific parameters such as, for uration specification, these included: hasInstalledSoft- instanxce, minimum and maximum length of contract ware, hasOperatingSystem, isRunningOS ; and some for agent joining a team. Therefore, the ontology additional ones that link the original CGO classes but should be extended with means of defining constraints were missing from the ontology: hasStorageInterface, such as minimum, maximum value for continuous hasStorageSpace, hasFileSystem.
Moreover, the in the CGO. These properties include: User may want to specify weight of a parameter i. In the remainder of this section we present to execute a job and the LMaster offering to run this these modules in some detail. To deal with these two cases, the ontology intro- duces two new classes: JobExecutionConditions and 4. Let CGO that adds additional classes and properties that us now briefly describe classes comprising this ontol- make specifying the hardware and software configu- ogy and their properties.
Note that some of these classes and properties have been already defined 4. The proposed extension also specifies a num- time during which the job should be executed ber of properties that join concepts of the CoreGRID ontology.
Con- to the LMaster as a question about the terms tent of messages is based on terminology defined in the on which it might be accepted; it utilizes the of- aforementioned ontologies. At the moment we are in feredResource property and contains an instance the process of evaluating different approaches to on- of the GridComponent class which describes the tological description of parameter constraints such as configuration of the resource it is offering; it will these described in section 3.
Therefore, in the follow- also contain a definition of constraints on prop- ing list of classes from the AiG Messages Ontology we erties from the WorkerContractConditions class will omit concepts related to the constraint definition. Dominiak, M. Ganzha, M. Gawinecki, W. Ku- ranowski, M. The consumer may be a lated to a machine. In other words, it is necessary the group. However, it provides a producer first establishment some rules as a standard to clear understanding of characteristics and properties of allow access to resources.
The next step is characterized classes and relations. In addition to that, it is possible to by a service level agreement between a producer and a extent for the use of new domains. New classes, rules or consumer of the services. This agreement does not reflect vocabulary can be added for a new application domain.
Ontology can be distributed and shared for utilization. In Gang-Matching [8] - in this research work is presented a conjunction with other ontology and tools can provide multilateral matching based on the consumer utilization also interoperability. Therefore, service oriented grid architecture requires a 2.
Ontology components scheme to support the interoperability between applica- tions from the VOs and a high level of access transpar- The representation of an existing element in the ontol- ency for resources. These components 4. Ontology for grid resources description are able to identify each category of elements in the on- tology. Elements and their meaning can have different In this section we present our research work that, dif- terminology based upon the area which utilizes these ferent from [1] [18] presented in section 5, targets the components.
However, in [6] there is an identification of construction of ontology for grid resources description. The components are: figuration. The target community was people from com- putational science.
As examples we can consider sub-class ontology requirements. Service-oriented in grid architecture time it is expected to allow access to a specific re- source; address of the resource; capacity of storage; In [2] it is argued that the concept of a grid is moti- available memory; existing operating system; archi- vated by a real and specific problem. The authors believe tecture type. These environments are source. Thus, every time that a request comes to a called as virtual organizations VOs.
VOs are character- semantic view, the structure returns information ized by a number of organizations or individuals provid- about that moment. In other words, a semantic view ing and consuming resources and services from the grid, replies if the resource is available for use, out-of- following rules of sharing.
Therefore, a large portion of work or heavily loaded between others status. Processors, memories, disks and software packages Metadata and semantics view are both used as addi- are able to be accessed from any node of the grid. How- tional references to ontology. Even not commonly used in ever, to reach an ideal access transparency for resources other ontologies, these structures can improve the ontol- and services it is necessary a complex control of produc- ogy action, returning answers more quickly.
Because semantics figuration. The ontology component utilizes the meta- tures have updated information of computational re- data and the semantics view to obtain information related sources. The metadata and semantics The ontology developed in this work was implemented views components help the ontology to obtain answers to using the OWL Language Web Ontology Language.
The metadata structure receives in- This language is used as a standard by the W3C. The formation directly from resources and data files. Figure 1 illus- Schema.
OWL is designed for use by applications that trates a data flow necessary to obtain information using need to process the content of information instead of only the ontology approach in a grid environment. This feature provides mechanisms to create properties and classes, allowing the creation of relationship between classes, cardinality and characteris- tics of properties. These three languages differ from each other in the level of formality provide to users to create the ontology.
In this research work we use the OWL Full. This language Figure 1. Grid architecture using the on- has an interesting level of formality and freedom, these tology approach. Methodology cies from axioms. In this environment, we describe con- establish which methodology is more suitable for the en- cepts, attributes and relationships of the ontology.
We vironment. Therefore, we consider the portability an es- chose this editor because it is free available, has support sential factor to be considered. Enabling Knowledge Discovery Services on Grids. European Across Grids Conference. Data mining ontology development for high user usability.
Wuhan University Journal of Natural Sciences. This paper mainly introduces the development and implementation of the user-centered data mining service ontology on Universal Knowledge Grid UKG , UKG is an ontology-based grid architecture model … Expand. Distributed data mining on grids: services, tools, and applications.
Computer Science, Medicine. View 1 excerpt, cites background. An ontology reasoning architecture for data mining knowledge management. Cooperative Inf. IJCAI Tangmunarunkit, H. Ontology-based resource matching in the grid—the grid meets the semantic web.
In Proceedings of 1st workshop of semantics in peer to peer and grid computing in conjunction with 12th W3C, Budapest. Corcho O. Journal of Web Semantics 4 2 : — De Roure D. R, Shadbolt N. Wiley, USA, pp — Vidal, A.
Wieder, P. Bringing knowledge to middleware—grid scheduling ontology. Satyanarayanan, M. Pervasive computing vision and challenges. Download references. You can also search for this author in PubMed Google Scholar.
0コメント