Guide to Metadata for Science and Research
Distributed Informational Retrieval for Science
Semantic Web for AURIS-MM project
Topics of interest:
Andrei Lopatenko's Projects
- Tools for developers of Semantic Web solutions. ETL Tools for Semantic Web
- AURIS-MM. Semantic Web for CRIS
- Innovation projects management system for Russian Fund of Basic Research (RFBR)
- Workflow and information resources life-cycles support for Scientific Digital Libraries
- Document Management System for Scientific Secretaries of RAS, research project management system for committees of RAS
Tools for developers of Semantic Web solutions. Extraction-Transformation-Load Tools for Semantic Web (Jan 2000 - )
The objectives of the project are
 |
to develop environment for development Semantic Web solutions. These tools must embrace full life-cycle of development. Conceptual analysis, ontologies, schemas, application profiles, vocabularies, transformation rules. |
|
to develop toolset which allows to connect traditional RDBMS solutions, Object-Oriented solutions (Enterprise Java Beans, etc) and XML with Semantic Web. |
|
to implement a library of asset which can help to develop solutions based on Semantic Web for research and technology |
Developed application for creating and supporting application profiles, RDF Schemas, controlled vocabularies, thesaurus mappings. The application make it possible organize simple life-cycles of the development semantic web solutions. OIL ontologies can be loaded and user in development metadata schemas. Mapping from OIL ontology to metadata schemas (RDF), to database table, view, procedure definitions, to classes (UML) can be established and browsed. Experience shows that it can help much in development Semantic Web solutions. Also mapping between RDF Schema elements and UML classes or tables can be inputted and browsed for development of RDF Schemas based on already developed systems or specifications, or for RDF publishing/loading/using.
Developed MOF (Object Management Group (OMG) Meta Object Facility ) and CWM (OMG Common Warehouse Model) compatible in data structure repository for storing information about classes, tables of the information system. Developed expanded MOF repository for holding RDF Schema and Ontology information
Developed the application for ETL (extraction-transformation-load) information operations. Using the application, schemas of data source and target can be defined (RDF Schemas, RDBMS schemas, UML Model). Also vocabularies used in data source and target can be defined. The application allows describe data mappings or data transformation process between data source and target (description is based on OMG Common Warehouse Model). Class mapping, attribute mapping, vocabulary transformation and more complicated operations can be defined. Then application performs ETL operations. The data can be exported and loaded from RDBMS or Object-Oriented system such as EJB into RDF/XML. The application uses ontology and schema information to solve semantic and structure interoperability problems
Using the tools resulted from the project, developed RDF Schema for CERIF and tools for exporting, loading data from CERIF database to/from RDF.
Currently the ontology for CERIF is being developed, mapping from CERIF RDF Schemas into other formats for research&technology&education(Math-NET metadata, KA2 Ontobroker, OAI ) is being developed
May 2001 - .Vienna Austria. Semantic Web for AURIS-MM
Solution for research data exchange based on semantic descriptions in RDF, vocabulary descriptions in NKOS () framework and OIL were tuned for Austrian and European standards. User interfaces for description of vocabularies and RDF schema were developed and improved.
OIL Object Inference Language language, based on RDF for description of ontologies, semantics, thesauruses. The Ontology Inference Layer OIL is a proposal for a web-based representation and inference layer for ontologies, which combines the widely used modelling primitives from frame-based languages with the formal semantics and reasoning services provided by description logics. It is compatible with RDF Schema (RDFS), and includes a precise semantics for describing term meanings (and thus also for describing implied information).
(www.ontoknowledge.org/oil/)
NKOS (Networked Knowledge Organization Systems)- organization related to the usage of knowledge organization systems in Internet based services and digital libraries. These systems can comprise thesauri and other controlled lists of keywords, ontologies, classification systems, taxonomies, clustering approaches, dictionaries, lexical databases, concept maps/spaces, semantic road maps etc.
http://www.alexandria.ucsb.edu/~lhill/nkos
Conference
For EuroCRIS conference all work was documented. PowerPoint presentation was prepared PowerPoint presentation, which consists of more 50 slides and describes
- Requirements for data integration solutions
- Application profiles framework for development of application-specific metadata schemas
- User interfaces which were developed
- Examples of using RDF for descriptions of CERIF data
- RDF data management facilities
Work was presented at Amsterdam EuroCRIS 9th platform meeting and it was discussed there. It was cause of interest of researchers
Installation
To run database, data integration utilities and develop application, a number of applications were installed. Among them are Oracle 8.1.7 database Enterprise Edition, Microsoft Visual Studio 6.0, Microsoft Visual Java 6.0, Oracle JDeveloper 3.1, Tomcat server 3.2.1, Apache Xerces (XML parser), OilEd tools for development ontologies on Ontology Inference Layer, Sun Java SDK 2EE, ActivePerl, Oracle Desgner 6.0. Also Russian variant of database for innovation projects was installed for testing propositions
Comparison of CERIF and AURIS
Works to develop data integration solution in Austria were begun. The comparison of CERIF and AURIS schema and vocabularies was done and documented. This document allow to easily develop data integration solution between AURIS and CERIF compatible databases. It describes relation between CERIF and AURIS tables and columns, description of compatibilities and incompatibilities. For each CERIF table and column in a base CERIF model, document describes in which AURIS tables and columns this information is stored and how it can be extracted
AURIS-MM Database
Database, based on CERIF for AURIS-MM project was developed. It was developed for Oracle 8.1.7 platform.
CERIF Oracle scripts were installed with error corrections. Database deals with such types of objects as persons, projects, organization units, relations between them. Some CERIF vocabularies for better description of Austrian data were substituted Austrian vocabularies. But data integration solution allows define vocabulary mappings from Austrian vocabulary to CERIF and reverse, so AURIS database can integrated with European automatically.
CERIF use European specific vocabularies for description of expertise, research area. They are very general and do not satisfy university community of Austria. OSTAT vocabulary was installed into CERIF database, which allows users to classify data by Austrian OSTAT vocabulary, which has a proved quality and lot of users experienced with it
AURIS-MM User Interfaces
To provide services for users, internet visual interfaces were developed for database. They were developed on Java Server Pages platform and were tested on Tomcat server. The database connection established trough jdbc. Already interfaces for browsing list of projects, persons, organization units were developed. Also interfaces for seeing information about projects, persons and organization were developed. These interfaces allow to see complete information about these objects and see relation with other objects. User interfaces allow navigate from object description to descriptions of related objects. Also interfaces for entering information about persons, projects and organization units, and relation between these objects were developed. JSP platform allows to easily use this application on a number of platforms, both commercial as well free, such as Apache, Oracle, Tomcat and others.
Correction of CERIF
While developing CERIF based database, a number of errors in CERIF implementation were found. This CERIF errors were corrected in a way to store general compatibility with CERIF both at semantic level and schematic level. The errors were documented as well our solutions to correct them. The description of errors is ready to discuss in EuroCRIS CERIF Task Force.
There were identified several CERIF errors, most of them are incompatibilities between check constraints and datatypes. This errors can be corrected by disabling Oracle constraints.
RDF generation
Development of RDF tools, generating descriptions of organization units, persons and projects in CERIF compatible were done. This tools were made compatible with internet jsp server what allows to publish RDF directly into web pages for semantic resource descriptions and for using of RDF by researchers.
Innovation projects management system for Russian Fund of Basic Research (RFBR)
Short description of the system in Russian
May 2000 - December 2000
Developed CRIS solution for dissemination R&T information about the results of the research projects funded by RFBR.
The application holds data about institutions, researchers and research results for technology. The main data are about research results, The data about about research results are linkd with data about institutions, persons, grant of RFBRs.
Information about research results can be entered by researcher or by experts of RFBR. Experts of RFBR can monitor information, check data, estimate data.
Search and browse service allows to find about R&T results searchng in database by attributes values, vocabularies or browsing vocabularies
The application is secure, administrator can define user accounts, assign them rules. Access to data is restricted according to rules defined by administrator
Workflow and information resources life-cycles support for Scientific Digital Libraries
Will be added
Will be added
.