Semantic Web-portal

Project № ІТ / 558 – 2009 July 20, 2009
“Semantic Web-portal of knowledge and intellectual analysis Development for the National Grid-infrastructure”
The project is dedicated to the National Grid education segment improving – the infrastructure for e-science needs, that was created on a previous stage of the State Program IKT implementation according to the Ugrid project № ІТ /506 - 2007 August 22,2007.It concerns the beginning of developing of the 3rd wave of Grid systems, so called Semantic Grid, where data and services have clear definition, which provides people and computers to work in cooperation. Semantic Grid as an extension of modern Grid will be really helpful for e-science, because its technologies will give an opportunity to work easy and quickly with knowledge not with information as it is being done now. Semantic Grid Applications can integrate a large number of heterogeneous information and service sources, which are involved and combined, as well as human, computational and transmission services.
Work Results:
1. A toolset for semantic support of HTML-documents, especially content management system (CMS), which has incorporated all today’s available micro formats and other semantic technologies was developed. While publishing the information through this system, a semantic formatted document can be automatically generated, and it will have more chances to be indexed after the search engines processing and it is located on the first pages of search reply.
2. Advantages and difficulties in technology choosing of web-services were analyzed as basics for the realization of Grid-service concept. The main contradiction between principles of service-oriented architecture and the realities of grid-environments were explored, also possible ways (already realized and perspective) of their overcoming were considered. Recommendations for Grid-service developers of the National Grid-infrastructure about the coordination of Grid- and Web-services were stated. The initial range of standards that are recommended for developer’s usage in problems minimization with compatibility of developed solutions was specified
3 The suitability of individual standards from the stack of web-services specifications (WS-*) and Web-services recourses (WSRF) for the implementation of grid-services was analyzed. Following aspects were considered: compatibility with existing tools, capability of grid-environment features, the ability to integrate into the security infrastructure, possibility of using standardized automated features of composite processes execution, semantic extension capabilities.
4. Grid infrastructure elements were analyzed to find those elements that can be improved with the help of semantic technologies. These elements are the most dependent from search operations, matchmaking and decision making: information system, brokers, workflows.
5. The methodology for creating ontology that was proposed is based on declarative knowledge presentation systems. Existing ontology description languages were analyzed and compared (OIL, DAML+OIL, OWL). The examples of ontology for Earth sciences and Grid-technology as a modern base for computer sciences were provided.
6. The software semantic search engine was created as a tool for browsing and searching in semantic data based portal. It consists of two components-server and client.
6.1 The server is presented with 2 services:
-SPARQL-processor for general purpose
• client’s request handler
SPARQL-processor implements a standard W3C protocol for remote SPARQL queries for RDF data sets.
It is built on the base of open Joseki code servlet.
Access Service for remote queries consists following components:
• Joseki servlet to implement the protocol 
-SDB-Jena framework library for working with semantic data, which allows to use relational data base as a storage for triplets.
-MySQL DBMS
Request handler is implemented as a servlet that takes query type and retrieve object as input, conducts its validation, performs SPARQL-query to the repository and returns results in JSON format for displaying in client’s component.
6.2 The client component is a JSP-page with built-in scripts (JavaScript), using jQuery library with plug-in for displaying and dynamically updating the page content based on the results returned by server component.
6.3. JSP-page (index jsp) allows performing key word, data features searches and displaying a list of resources, that meet the case. In addition, with reference transitions in returned search results all features are displayed that are connected with this resource that is currently present in semantic basis, in tabular format.
6.4 Advanced.jsp page provides the query constructor to a semantic data base with an opportunity for describing random number of statements that have to be satisfied with project’s results, rules for result filtration, their sorting and limitation of their amount. Every statement has a subject, predicate and an object. Queries results are displayed in the table. Each row has a number of columns, that fits the number of variables in the query (variables are marked).
Resources that are received as a query result are the links for viewing related features as it is described above.
7. A tutorial “Semantic Grid for science and education” (Petrenko A.I., Bulah B.V., Hondar V.D.) was prepared. It supports the new course of Grid-technologies for Universities, its teaching will start for masters in February 2011 in NTUU “KPI”.
The practical significance of results:
Nowadays The Development and implementation of Semantic Web technologies belong to the main-line direction of web-technologies progress. Usage of Semantic Web technologies in educational portals allows providing higher level of educational personalization, to increase quality of processing user’s requests, to integrate more efficiently information from heterogeneous sources for benefit of scientific and educational activity.
Thereby the development of semantic portals has to be considered at this stage as a priority for creation of educational and research information environments of the next generation.

Assignment’s actuality is confirmed with the increase of search engines’ interest in semantic formatted pages, with constant introduction of new micro format and improvement of existing ones. One can but consider that all the major services (Facebook, Twitter, YouTube etc) have already been using numerous micro formats for search return improving and for creation of new connections between people and events. Search giants ((Google, Yahoo, Yandex etc) every day monitor semantics’ development and service development in particular, to insure further improvement of their searching mechanism.
Analytical generalized Semantic Web materials are used for developing program and information for new discipline- “Web-design and Web-technologies”, that was recently introduced in the curriculum for bachelors with “Computer Science” major according to the State Standard Project for education.
A study program for master’s course “Grid technologies for distributed computations and data processing” (s.s.s. Svistynov S.Ya.) was developed, it has “Portals and meta planners” section. 
Analyses of possible architecture and features of Semantic-Grid that were accomplished, provide foundation for qualified technical assignment generation for competitive projects(according to the new State Program for Grid-technology implementation for 2009-2013, particularly for its section on “development and execution methods of efficient usage of existing Grid-resources through Data Mining”).

11 articles and results of the project on the subject of Semantic Grid were published, 10 presentations at national and international conferences with thesis publication were performed. Educational materials “Semantic Grid technology for science and education” were published as a part of the new university course -“Grid-technologies” which will begin in February 2011 in NTUU “KPI”.
Results documentation:
1. Project site www.portal.hpcc.kpi.ua which is a semantic Web-search engine with data, particularly for grid-resources data. 
2. Site for distance learning- http://moodle.kpi.ua,that includes educational materials for remote masters’ course “Grid technologies for distributed computations and data processing”


Ukrainian (UA)Russian (CIS)English (United Kingdom)