Periodic Activity Report (2008)
ESTRELLA brings together the leading European companies and experts in the market of legal knowledge systems, to collaborate on the development, demonstration and dissemination of an innovative, open platform enabling citizens and businesses to easily access, understand and apply complex legislation and regulations. The interests of public administration are represented in ESTRELLA by the participation of the national tax ministries of Hungary and Italy.
The main technical objectives of the ESTRELLA project were to develop a Legal Knowledge Interchange Format (LKIF), building upon emerging XML-based standards of the Semantic Web, including RDF and OWL. ESTRELLA also designed a document management system for legal sources, based on a uniform XML format, enabling search, explicating relations with other documents and LKIF models of their content, and version management.
To demonstrate and validate the ESTRELLA platform, European legislation and national tax legislation of two European countries were modelled and used for pilot applications.
In the first year of the project a first version of LKIF was specified. In this second year, LKIF was tested in actual use in three pilots and already refined. In this third and last year the final version of LKIF was produced and published. The expressiveness of the language increased, the integration with the CEN MetaLex XML format for sources of law was improved and LKIF is easier to extend now.
The reference inference engine for LKIF – implemented in Scheme - that was produced (named ‘Carneades’), the API’s for interacting with LKIF knowledge bases, and the translators from and to the vendor formats were improved. Carneades has been ported to the R6RS scheme standard and now runs on 3 different open source scheme implementations. It now supports complex rule bodies and has a built in ontology reasoner for the DLP dialect of description logic. It also has a prototype graphical user interface. The API now provides a programming language-independent, web-oriented interface to the inference engine based on SOAP and WSDL.
Additionally, the project experimented with a normative assessment (a central subtask in all legal reasoning) approach entirely based on OWL DL. This lead to the implementation of HARNESS (“Hybrid Architecture for Reasoning with Norms Exploiting Semantic web Standards”), a Protegé plugin that reasons with norms expressed in OWL DL 2.0. The consortium has also contributed to the specification of OWL DL 2.0 and participates in the W3C OWL working group.
Concerning the management of legal sources, the project has continued co-operation with the CEN/ISSS standardization activities initiated by UvA early in 2006. The project contributed significantly to the new version of CEN MetaLex that was adopted in the September 2008 meeting in Florence (Italy). The translators between national standards and CEN MetaLex were improved and adapted to the last version of CEN MetaLex.
An interchange format for metadata of sources of law was developed, called LMIF (Legal Metadata Interchange Format), that enables translation of national metadata and CEN MetaLex and coupling to LKIF models and ontology.
An Open Source legal CMS based on CEN MetaLex, called eXistrella, has been designed and delivered.
The ESTRELLA business model and Exploitation Plans have been delivered. Specific exploitation plans are declared by the industrial partners and by the research groups within the project. A key part of the exploitation focus is the promotion of LKIF as a CEN Standard through the formation of a working group. This starts with the ESTRELLA partners and will grow to include all interested parties requiring representation in that activity. Besides Allex Gold (Hungary), two small Italian companies and a Dutch company in the field of legal KBS have shown concrete interest in complying to LKIF.
Many scientific papers on ESTRELLA results were produced in the period.