Design and Implementation of a Prototype for Semi-automated Determination of Jurisdiction and Applicable Law

Embedded in a larger research effort towards an automated determination of jurisdiction and applicable law for service contracts in the Internet, this thesis adopts a previously developed PIL (Private International Law) modeling methodology as its starting point. Based on that methodology, the first cornerstone in this thesis consists in extending the range of modeled (national and supra-national) PILs considerably. Currently, there are models available for the Brussels regime which reflects questions of jurisdiction and applicable law on a European level, plus there are models reflecting the Swiss national perspective. In this context, considerable extension of modeled PIL range means to model two further national or a single supra-national PIL by applying the modeling procedure developed and used previously. Those numbers indicated are understood as minimum numbers. This model effort may lead to a required adaption of the methodology. If this is the case, the accordingly altered methodology is documented in this thesis and the respective changes needed in previously modeled PILs are listed, whereas existing models need not to be updated within the frame of this thesis.

Subsequent to that extension of modeled PIL range, this thesis is concerned in terms of a second area of work with the design of a prototype for (semi-automated) determination of jurisdiction and applicable law in relation to international contracts of civil and commercial matters, such as for an international contract of a commercially provided electronic value-added service in the Internet. This design task is expected to account for the primary work load caused within the scope of this thesis. This is mainly due to its complexity, which, in turn, originates from challenges to be expected in integrating multiple PILs in a single design. Major integration challenges are found accordingly in the design of a consistent information and workflow model. The information model needs to reflect, on the one hand, specifics of a single PIL—e.g., the respective definition applicable of habitual residence—while it needs to base on commonly derived notions on the other hand. In other terms, the information model needs to model information objects, such as connecting factors, in a technically correct way that does not inflict with any of the modeled PIL’s respective notions. As with regard to the workflow model, challenges are to be expected in the question of when to apply which PIL. This relates to determining the set of conditions and rules of prevalence that indicate a legal domain is probably touched by a considered international service contract. In addition to a consistent information and workflow model, the design task is concerned with the prototype’s architecture. This incorporates design decisions in the areas of software components to be developed as well as the interfaces and interaction between these components. Technology-wise, the implementation design is not bound in any way by a dependency path. However, a distributed systems design is a must, plus an implementation in Java is proposed.

The third cornerstone within the scope of this thesis bases on the respective results of the first two, those of an extended range of modeled PILs and of the design task. It addresses an implementation in terms of a prototype and an evaluation in terms of functionality validation. The prototype implementation shall simulate a semi-automated determination of jurisdiction and applicable law in relation to an international service contract to be concluded between a service provider and a service customer. Semi-automated refers to this prototype’s execution and configuration modalities. A single prototype execution run is expected to cover parts of a business transaction, namely that transaction part to address PIL-relevant issues in contract negotiation. In order to cover this transaction part, the prototype takes a set of configurations as input. These configurations reflect the relevant set of connecting factors from the involved service provider and service customer, respectively, as well as from the service specifics. Hence, these configurations have to be pre-compiled so that they are not part of a fully automated determination. Full automation, however, is expected from the prototype’s execution in terms of a single test run. In other terms, once a test run is set up—i.e., the relevant configurations are made available to the prototype—the prototype is executed and the respective result set is determined automatically. A result set may be either an exception caught (in case of prototype execution exception) or a set of suitable jurisdictions and applicable laws (in case of regular prototype execution). Once results are available from a complete test run, obtained results are compared with the respective expected test result documented time-wise before the actual test run in the prototype. Functionality validation, hence, is seen as the core outcome of this implementation and testing task. Every test run in the prototype shall reflect a previously documented test case. The set of test cases is expected to cover a wide range of connecting factor combinations. For every test case—each test case finds technical representation in the according set of prototype input configurations as well as the result set obtained—, this thesis needs to document whether the prototype is able to provide the expected functionality.