Modeling a Semantic Web management system.

Carstoiu, Dorin ; Cernian, Alexandra ; Spanoae, Gavrila 等

1. INTRODUCTION

The Web was designed as an information space, which should be useful not only for human-human communication; machines should be able to participate and help. One of the major obstacles to achieving this goal has been the fact that most information on the Web is designed for human consumption. Even if a certain company has got a system for finance management, another system for human resources and perhaps a separate one for process management, even if the information is derived from a database with well defined meanings for its columns, the structure of the data is not obvious for a machine. Even for an employer it is difficult to find out where information in located. The Semantic Web approach develops languages for expressing information in a form that can be processed by a machine (Buraga, 1998).

Metadata (meta data, or sometimes meta information) is "data about other data" of any sort. An item of metadata may describe an individual content item or a collection of data including multiple content items and hierarchical levels, for example a database schema. In data processing, metadata is data that provides information about the data managed within an application.

When structured into a hierarchical arrangement, metadata is more properly called an ontology or schema. Both terms describe "what exists" for some purpose. For instance, the arrangement of subject headings in a library catalog serves not only as a guide to finding books on a particular subject in the stacks, but also as a guide to what subjects "exist" in the library's own ontology and how more specialized topics are related to or derived from the more general subject headings. Metadata is frequently stored in a central location and used to help organizations standardize their data.

Therefore, if we define a system which combines the benefits of the Resource Description Framework (RDF) with database metadata, we will get better performances in retrieving information in real time.

The Semantic Web Management System (SWMS) we propose in this paper enables the users to achieve their goals by transforming disparate, fragmented data into viable information capable of answering key questions. Data from across the web, enterprise, business units and functional areas is reorganized and recombined to report key information about information (such as: where is the useful information). It provides real-time analysis capabilities that increase "the speed to insight". The SWMS architecture is far more efficient in the deployment process than a common application.

2. RELATED WORK

We have found different interesting approaches for creating an infrastructure for the Semantic Web in (Dokulil et al., 2008) and (Vitvar et al., 2007). In (Dokulil et al., 2008), the authors propose Trisolda, a Semantic Web infrastructure built around a semantic repository with importing, querying and data processing interfaces. The authors also propose the TriQ RDF query language, designed especially for complex semantic querying. In (Vitvar et al., 2007), the authors present the WSMX environment, concerned with semantic web services. Its purpose is to allow, mediation, invocation and inter-operation of the services.

Our approach is more closely related to the Trisolda approach, both of them creating a semantic repository where RDF files are stored. In fact, storing RDF files into a database is not a novelty and there are already many systems which allow this sort of operation. The novelty we propose is to create a meta-model of the RDF files stored in the repository. Thus, we generate a map of the information contained in these files, which will improve the performances when browsing through the data. This meta-model will also be stored in the database.

3. SYSTEM ARCHITECTURE

From an architectural point of view, the Semantic Web Management System (SWMS) is composed of different modules working together.

The main goal of this concept is to integrate the Semantic Web with the actual needs of a research, development or even corporate process.

The architecture consists of three main layers:

* a Consumer Layer which interfaces the resources (the RDF source files)

* a Middle Layer which holds certain resources. This layer is the mediator which enables a common communication language between the source of information and the SWMS

* the Warehousing Component which stores and prepares the information for future browsing or updates.

In our architecture, the Warehousing Component and the Consumer Layer are permanently connected and certain jobs are executed on scheduled program to ensure an automated process.

The system has the following components:

A. Importing Interface

This sub-component interprets the input files and once the RDF format is identified, the corresponding tables in the database are created.

B. Data Integration

Here, the information is integrated in the Semantic Meta Model and in the Semantic data store. For this operation, the system uses a number of views, such as the Data Access View from the database Importing Interface layer, which based on triggering events and data composition knows how to integrate the information in the database.

C. Data Access Layer

This sub-component is the communication bridge between all the application interfaces and database layers. It is also the component which helps the SWMS application to communicate with the end- user.

D. Query Interface

The Query Interface component works in collaboration with the Application Interface Semantic Meta Model layer of the database in order to facilitate the management of data. The user will be able to easily launch complex queries based on SPARQL in order to satisfy their reporting needs.

E. Data Export Interface

When a set of RDF files are imported into the database, with the help of the Semantic Meta Model and the Exporting Interface layers, a third party RDF is created in order to combine all the information contained in the ontologies from the data store in one big ontology RDF file which could be exported and used by a another semantic application or even published on the web.

F. SWMS Database

The Database contains different layers of Data Access:

* Importing Interface--It contains a number of Data Access Views that, based on triggering evens and/or business logic, integrate the RDF files information in the database;

* Semantic Meta Model--It is a generic database model. Here, information about the data that is stored in the data store section is recorded, together with information regarding user enrolments and reports definitions.

* Semantic data store--This is the most dynamic part of the system because it grows with any new file that is imported. For each file in the data store a minimum number of tables is required: a table that replicates the file in the database, a table that represents the RDF entities and another table linking the entities in order to define the ontology graph or even parallel graphs.

G. SWMS Application

The Semantic Web Management System Application is a simple and user friendly application, which at present is in the design phase. It will be made of 6 screens: Main Menu Console, Admin Console, Importing Console, Reporting Console, Exporting Console and Help Window. The application has got a simple design and it is using the open source libraries for Jena Semantic Web Framework, but also other open source and in house developed libraries.

The scenario for which this system was conceived is the following: In a company, each department is using dedicated software applications in order to manage data and documents. Nevertheless, some pieces of information might be interconnected. For a certain task, somebody may need to know exactly where to find specific information, as well as the links and relationship between different information entities. Without a centralized and efficient repository, this is rather difficult to accomplish.

[FIGURE 1 OMITTED]

4. CONCLUSION

Semantic technologies are being added to enterprise solutions to accommodate new techniques for discovering relationships across different database, business applications and Web services. RDF triples are persistent, indexed, and queried, similar to other object-relational data types.

To conclude, the benefits of combining semantic web concepts and database metadata in one Information Management Tool are: improved data access, scalability and platform independence.

As future work, we plan to implement and validate the architecture presented above.

5. REFERENCES

Buraga, C. (1998), An Advanced Concurrent Teleconferencing System, Proceedings of the 6th International Symposium on Automatic Control and Computer Science--SACCS'98, Bucharest, Romania.

Dokulil, J., Yaghob, J and Zagoral, F. (2008). Trisolda: The Environment for Semantic Data Processing, International Journal On Advances in Software, vol 1 no 1, pp. 43-58.

Vitvar, T, Mocan, A., Kerrigan, M., Zaremba, M., Moran, M., Cimpian, E., Haselwanter, T. and Fensel, D. (2007) Semantically-enabled service oriented architecure: Concepts, techology and application, Journal of Service Oriented Computing and Applications.

***W3C Semantic Web, http://www.w3.org/2001/sw/, Accesed on: 2009-06-01

*** RDF, http://www.w3.org/RDF/, Accesed on: 2009-06-01

***SPARQL, http://www.w3.org/TR/rdf-sparql-query/, Accessed on: 2009-06-01