Workshop Jurix 2012

Contents

Acquisition and management of legal information in the semantic web

With the maturing of information exchange over the internet, many organizations face the challenge to seamlessly exchange information with a large and changing group of customers as part of their business processes. Providing and using information cannot be done without the correct interpretation of that piece of information. An open way of publishing, providing and acquiring of these semantics is needed.

Governmental organizations in particular face the challenge to respond to the needs of citizens and organizations to transparently supply information that can be interpreted correctly by citizens and organizations, by providing semantics that are easy to read and at the same time comply to law and regulations.

One of the cases that will be presented during the workshop was executed at the Dutch Land Registry Office, Kadaster, that has the legal obligation to collect, record and provide information about registered properties in the Netherlands. In this case study, academic research on legal knowledge acquisition is combined with semantic web technology and practical experience, resulting in a manageable and transparent way to annotate information products with semantic tags that are not only understandable for the customers of the Kadaster organization, but can be traced back to its legal roots as well.

Two papers are accepted:

The Dutch Chamber of Commerce contributed a case.

A small introduction to the workshop is available as a prezi.

Registration for the workshop

The workshop is part of the Jurix 2012 conference in Amsterdam. The date of the workshop is december 17th 2012.

Registration for the workshop is possible via the Conference webpage.

Outline of the workshop

The workshop aims to initiate an interaction between the members of the academic research field of legal informatics and organizations in legal practice (in particular governmental organizations) to improve the method that was used the Kadaster case study. The results of the workshop will be published and integrated with the SOM approach described on this wiki.


9.00-9.20 Registration

9.20-9.30 Plenary opening


9.30-10.00 Workshop introduction (Marco Brattinga)

10.00-10.30 Standardizing for Mining Complex Legal Documentation (Nada Mimouni)

10.30-11.00 Organization and retrieval of law sources by utilizing technologies of the semantic web (Marios Koniaris)


11.00-11.30 Coffee break


11.30-13.00 Introduction Semantic Open data Method and case Dutch Cadastre (Marco Brattinga, Arjen Santema, Miriam van Dalen)


13.00-14.00 Lunch


14.00-14.15 Introduction case Dutch Chamber of Commerce (Rob Onink)

14.15-14.45 Hands-on workshop acquisition and management of legal information (Rob Onink, Marco Brattinga, Arjen Santema, Miriam van Dalen)


14.45-15.15 Coffee break


15.15-15.45 Continuation hands-on workshop

15.45-16.45 Discussion to improve the methods that were used in the case studies, to create a method that can evolve into a standard method for semantically annotate information models within the context of the law.

Submission process and workshop attendees

Organizations and members of the academic research field are invited to submit a brief description of their case via email to marco.brattinga@ordina.nl. Submission deadline is December 3th, 2012. The description should describe the problem area and the method used, in the same way as the Kadaster case is described on this page.

The workshop organizers look forward to discuss their ideas with members of organizations and researchers within the same field of interest. They have already discussed some of the results of the case study within their network of professionals at different Dutch governmental organizations, and will invite these professionals to continue the discussions at the workshop.

The workshop organizers are:

Introduction to the topic - the Kadaster case

The Dutch Land Registry Office, "Kadaster", has the legal obligation to collect, record and provide information about registered properties. Almost all information products of the organization are offered in a digital way to customers, and an increasing amount of legal documents are collected digitally.

The Dutch land register was introduced in 1832 as a system of deeds derived from French civil code. In such a system, a Registrar files copies of legal documents in the Land register. The Registrar does not determine who is the legal titleholder. In 1929, the Registrar was also made responsible for the administrative part of the Kadaster. This resulted in a significant development of the system in the direction of the other known system in Europe, a title system. Nowadays, the Kadaster administration does maintain a registration of (presumed) titleholders, although from a strictly legal point of view, it is still a system of deeds.

The Kadaster administration is part of a federated system of key registrations within the Dutch government. Every individual key registration focuses on a part of Dutch civilization, like the register of properties, the civil register and the trade register. Governmental organizations depend largely on information within these registrations. From a legal point of view, these key registrations are the only valid and authentic sources of information for a particular topic. All governmental organizations have the legal obligation to use these key registrations. At this moment however, no integrated semantic model is available. This is due to the fact that different governmental organizations have to comply to different laws and information in one system can have different semantics to different organizations. No clear solution is yet available, resulting in a loss of information, higher costs to implement policies and the possibility of fraud.

Solutions are in place to deal with technical and syntactical interoperability issues. Electronic exchange formats are used and described using open standards as XML and UML. With an increasing and rapidly changing group of customers, semantical problems are more and more the main reason for interoperability issues.

In this case study a solution is investigated in which exchange formats are annotated with semantics and linked to parts of law and regulations. Every piece of information is identified using the publicly available data model of the Kadaster by giving it a unique URI. The semantic annotation is provided by a framework based on RDF/OWL and the academic research of Breaux[1] and van Kralingen[2]. Relevant law and regulations are processed using the Metalex standard[3]. In this way a link can be made from the semantic annotation to the law and regulations by referencing the URI via the Metalex standard.

By using an open-linked data approach, it is possible for the Kadaster organization to publish its semantic models without the risk that these models conflict with models of other governmental organizations. Semantic reconciliation between organizations as part of the information exchange is a real possibility. By encouraging other registration-holders to follow the same approach, an integrated and federated semantic model for the key registrations of the Dutch government is within reach.

Interoperability of information exchange is until recently somewhat of a IT driven area. We propose to shift the focus to the legal perspective. Information will still be the starting point, but the acquisition of the semantics will be done entirely by legal officers, and not by the IT department. Information technology can be used to give these legal officers the necessary concepts and tools to manage their work.

It is our experience that a top-down approach to the problem of legal knowledge acquisition is very cumbersome. It takes a very long time to formalize legal knowledge to the point that it can be used at the operational level. From an earlier effort it is estimated that a full elaboration of relevant law for the Kadaster would take at least two years. We propose a different approach with the information model as a starting point of reference.

Starting with the elements of the information model itself (entities, attributes), we obtained the candidate concepts for semantic annotation. The second step was analyzing the relevant law to acquire the events that the Kadaster organization should respond to according to law. A conceptual model for these concepts and events was created by a unique co-production of legal officers and IT specialists.

With this conceptual model as a framework, Kadaster employees with basic legal knowledge were able to annotate all the relevant concepts. A quality process was implemented to validate the correctness of the annotations by a registrar (the policy officer at Kadaster) and to verify the soundness of the model by a knowledge architect. As a semantic model of legal concepts, the resulting work can be used as a reference point to formalize the meaning of information products distributed by the Kadaster.

All results were processed using a prototype tool in which excel spreadsheets are imported and OWL/RDF open linked data is published. The prototype also has a web interface and search facility to facilitate publication of the semantics to end users and customers of the Kadaster organization.

References

  1. 1 Legal Requirements Acquisition for the Specification of Legally Compliant Information Systems, Travis D. Breaux. Ph. D. Thesis, North Carolina State University, 2009
  2. 2 A Conceptual Frame-based Ontology for the Law, R. van Kralingen. Proceedings of the First International Workshop on Legal Ontologies, 1997
  3. 3 CEN Metalex Workshop Agreement, Technical paper, 2010