A model-based approach to the automatic revision of secondary legislation

Abstract

Conflicts between laws can readily arise in situations governed by different laws, a case in point being when the context of an inferior law (or set of regulations) is altered through revision of a superior law. Being able to detect these conflicts automatically and resolve them, for example by proposing revisions to one of the modelled laws or policies, would be highly beneficial for legislators, legal departments of organizations or anybody having to incorporate legal requirements into their own procedures. In this paper we present a model based approach for detecting and finding legal conflicts through a combination of a formal model of legal specifications and a computational model based on answer set programming and inductive logic programming. Given specific scenarios (descriptions of courses of action), our model-based approach can automatically detect whether these scenarios could lead to contradictory outcomes in the different legal specifications. Using these conflicts as use cases, we apply inductive logic programming (ILP) to learn revisions to the legal component that is the source of the conflict. We illustrate our approach using a case-study where a university has to change its studentship programme after the government brings in new immigration regulations .

Contextualized Institutions in Virtual Organizations

“…We believe that such approach can actually resolve the conflicts by tracking and fixing the origins of them, rather than simply avoiding them. This approach to conflict resolution has been successfully applied to legal conflicts between cooperating legal systems [10] and because of the similarity, at the technical level, to the circumstances described here, we believe the same solution may be applied, perhaps save some minor details. By viewing each institution in a collective set (i.e.…”

“…with overlapping governance scopes) as a participating legal specification -to use the terminology of our earlier work [10] -the procedure is able to compute automatically all possible revisions of the existing norms in the light of the detected conflicts. In order to keep this paper self-contained, we provide a brief introduction to the conflict resolution approach, but for more details of the approach, please refer to [10].…”

Social Coordination Frameworks for Social Technical Systems

“…As a result, an actor could (unintentionally) be in an untenable situation where an action is permitted by one policy and not by another (defined as weak conflict), or obliged by one and not permitted by another (strong conflict). The first scenario [62] demonstrates the automatic conflict detection between two independent policies in which the first stipulates an upper bound on the working hours for the holder of a student visa, while the second sets a minimum number of hours that the holder of a studentship shall work. By constructing a composite trace for the two models, the constituents of the conflicting normative states are identified and then, by using the trace as a negative example, an inductive logic programming theory revision task is able to suggest a change that resolves the conflict.…”

“…[62] Examines the (automatic) detection of a policy conflict between two sets of regulation, where one (national) has precedence over the other (university) between legislation and the revision of the latter to comply with the former. …”

Policy-Carrying Data: A Step Towards Transparent Data Sharing

“…The situation is potentially further complicated since the platform may enable the collection and interpretation of those data, thus adding value to them, as in the case of activity-monitoring devices or home energy monitors. The PCD concept associates data with bespoke policies: for example, framework policies might be defined by legislation, while specific policies for individual needs would have to satisfy the norms established at the primary level [13] . In this way, crisp but unworkable definitions of issues such as "When do data stop being private?"…”