Contact Point:
Dr. Martin Doerr ( )
Institute of Computer Science
Foundation for Research and Technology – Hellas (FORTH) , Greece
  

Technical Contact Point:
Maria Theodoridou ( )
Institute of Computer Science
Foundation for Research and Technology – Hellas (FORTH) , Greece
  

Type of Software:
PROOF
  

Descriptive Keywords:
RDF knowledge base, CIDOC CRM, complementary information integration
  

General Description:
The VBI-ERAT-LVPA index is a knowledge base that integrates complementary archaeological information sources. The source data comprise complementary scientific databases and corpora describing finds with inscriptions and iconography of the Roman era. Integration is achieved by intellectually interpreting each source schema in terms of the CIDOC CRM model and storing it in an RDF knowledge base, thus creating a body of unique archaeological knowledge in digital form. Our main objective was to provide procedures for information extraction and global querying over all the contents of the complementary resources. Additionally we aimed at performing reliable statistical evaluation of the integrated data. In order to ensure that the methods used converge towards the best state of knowledge available and that the results are of high quality, we apply data cleaning procedures both at the individual sources and at the integrated knowledge base.
    

Technical Description:
The implementation of the system is based on three sets of tools. 
  
1. Transformation/mapping tools: They convert the data of various formats and sources into a common XML format compatible with the CIDOC Conceptual Reference Model. The input data come from the diverse archaeological sources and might be databases, text files, spreadsheet files etc. The results of the transformation are XML files, compatible with the CIDOC CRM. Most of the transformations are done using DataJunction 7.5 (http://www.pervasive.com/migrationtoolkit), an integration/migration tool, market leader in its field, designed to convert structured data from one format to another. It is also designed to clean and restructure the data to fit the new format. Additionally, Java and JavaCC (https://javacc.dev.java.net/) were also used for specific transformations, particularly for dictionary entries, such as OPEL. Finally, XML files are mapped to RDF descriptions through a specialized converter program. 
  
2. RDF Suite (http://athena.ics.forth.gr:9090/RDF/) which allows for effective and efficient management of large volumes of RDF descriptions. The produced RDF descriptions are validated with a Validating RDF Parser (VRP) and then loaded in the RDF Schema-Specific Data Base (RSSDB), a persistent RDF store that is used for the integrated knowledge repository.
  
3. SWPG, Semantic web portal generator (Athanasi 2004) in order to provide a Web-based easy to use by archaeologists user interface.

The user interface allows the formulation of three types of queries:

• Data cleaning queries, which will either produce reports that will be mailed to the respective source in order to improve and/or correct the data manually or activate procedures that will clean the integrated repository automatically.
• Queries of archaeological content to the integrated knowledge repository. With these queries we can detect contextual relationships that cannot be derived from interpreting the sources in isolation.
• Statistical queries.
    

Required User Skills:
No special skills needed.
  

Conditions of Use:
Contact the authors.
  

Additional Information: 

• Doerr M., Schaller K. and Theodoridou M. 2004. Integration of complementary archaeological sources. In Computer Applications and Quantitative Methods in Archaeology Conference (CAA 2004). Prato (Italy), 13-17 April 2004.
  (http://www.ics.forth.gr/isl/publications/paperlink/doerr3_caa2004.pdf)
• The CIDOC Conceptual Reference Model, Official Web Site http://cidoc.ics.forth.gr/