AI Approaches to the Complexity of Legal Systems

The paper examines how a particular class of robotic applications, i.e. service robots, or consumer robots, may affect current legal frameworks of privacy and data protection. More particularly, the focus is on (i) a new expectation of privacy brought about by these robotic applications; (ii) the realignment of the traditional distinction between data processors and data controllers; and, (iii) a novel set of challenges to the principle of privacy by design. Instead of a one-way movement of social evolution from technology to law, however, a key component of the analysis concerns the aim of the law to govern technological innovation as well as human and artificial behaviour through the regulatory tools of technology. Since most domestic and service robots are not a sort of “out-of-the-box” machine and moreover, their behaviour and decisions can be unpredictable and risky, special attention is drawn to the experiment of the Japanese government that has worked out a way to address (some of) the legal issues, which are at stake in this paper, through the creation of special zones for robotics empirical testing and development, namely a form of living lab, or Tokku. Interestingly, some EU member states have already followed suit.

The paper is an investigation on how behaviour relates to norms, i.e. how a certain conduct acquires meaning in institutional terms. The simplest mechanism determining this phenomenon is given by the ‘count-as’ relation, generally associated with constitutive rules, through which an agent has the legal capacity, via performing a certain action, to create, modify or destroy a certain institutional fact. In the analytic literature, however, the ‘count-as’ relation is mostly approached for its classificatory functions, mapping entities to categories whose members carry institutional properties. Besides making explicit this double function, the paper reconsiders the relation between constitutive rules and regulative rules, and introduces a proposal on the ontological status of constitution.

This paper discusses what types of truth play their role in law and in what way such types of truth can be explicated. For this purpose, this paper applies the logical viewpoint and method of predicate logic to clarify the logical structure of legal sentences and legal reasoning through their application. This paper presents its central argument that the concept of truth in law is to be classified into three types of truth, i.e., truth as fact, truth as validity and truth as justice, and provides their formal semantic foundation. This paper analyzes the way to explicate truth as validity and truth as justice in the ways of intensional and extensional explications on the one hand and in the way of the reasoning of justification and the reasoning of creation on the other hand.

One of the common characteristics of legal documents is the absolute preponderance of text and their specific domain language, whose complexity can result in impenetrability for those that have no legal expertise. In some experiments, visual communication has been introduced in legal documents to make their meaning clearer and more intelligible, whilst visualizations have also been automatically generated from semantically-enriched legal data. As part of an ongoing research that aims to create user-friendly privacy terms by integrating graphical elements and Semantic Web technologies, the process of creation and interpretation of visual legal concepts will be discussed. The analysis of current approaches to this subject represents the point of departure to propose an empirical methodology that is inspired by interaction and human-centered design practices.

Developing systems operating in alignment with norms is not a straightforward endeavour. Part of the problems derive from the suggestion that law concerns a system of norms, which, in abstract, in a fixed point in time, could be approached and expressed atemporally, but, when it is contextualized and applied, it deals with a continuous flow of events modifying the normative directives as well. The paper presents an alternative approach to some of these problems, exemplified by well-known deontic puzzles, by extending the Petri net notation, most common in process modeling, to Logic Programming Petri Nets. The resulting visual formalism represents in a integrated, yet distinct fashion, procedural and declarative aspects of the system under study, including normative ones.

The research for solutions for compliance is mainly focused on the representation of regulative rules, i.e. the imperatives that the industry is asked to comply to. Yet, a relevant part of the legal knowledge contained in regulation cannot be expressed in terms of deontic statements, and is instead represented as constitutive rules. This concept was first introduced by philosophers of language such as J.L. Austin and J.R. Searle and further developed in legal philosophy, where constitutive statements are classified in categories according to their legal effects. The present paper presents a heuristic approach for the representation of alethic statements as part of a methodology aimed at ensuring effective translation of the regulatory text into a machine-readable language. The approach is based on a classification of constitutive statements contained in the work of legal philosophers A.G. Conte and G. Carcaterra. The methodology includes an intermediate language, accompanied by an XML persistence model, and introduces a set of “legal concept patterns” to specifically represent the different constitutive statements. The paper identifies five patterns for the corresponding constitutive statements found in financial regulations: legal definitions, commencement rules, amendments, relative necessities, and party to the law statements.

Traditionally the integration of data from multiple sources is done on an ad-hoc basis for each analysis scenario and application. This is a solution that is inflexible, incurs high costs, leads to “silos” that prevent sharing data across different agencies or tasks, and is unable to cope with the modern environment, where workflows, tasks, and priorities frequently change. Operating within the Data to Decision Cooperative Research Centre (D2D CRC), the authors are currently involved in the Integrated Law Enforcement Project, which has the goal of developing a federated data platform that will enable the execution of integrated analytics on data accessed from different external and internal sources, thereby providing effective support to an investigator or analyst working to evaluate evidence and manage lines of inquiries in the investigation. Technical solutions should also operate ethically, in compliance with the law and subject to good governance principles.

This paper discusses the modeling of patent claims in ontology based representation of patent information. Our contributions relate to the internal structure of the claims and the use of the all-element rule for patent coverage. Starting from the general template for the structure of the claim, we present contributions to (1) visualization of claims, (2) storing claim information in a web semantics framework, and (3) evaluating claim coverage using Description Logic.

This paper describes a decision support system to represent the semantics of legal norms. The focus is on designing and software-technical implementing of a comprehensive system to support model based reasoning on legal norms and enabling end-users to create, maintain, and analyze semantic models, i.e. ontologies, representing structure and semantics of norms.A model based expression language (MxL) has been developed to coherently support the formalization of logical and arithmetical operations. MxL is intended to define complex, nested, strongly-typed, and functional operations. The paper summarizes research on the design and implementation of a legal expert system built upon model based decision structures. Thereby, three different components, namely a model store, a model execution component, and an interaction component have been developed. The formalization, execution, and analysis is shown on German child benefit regulations.

Legal causation is a complex aspect of legal reasoning. Due to its significant role in the attribution of legal responsibility, it is important that there is a clear understanding of the requirements for establishing and reasoning with causal links. This paper presents preliminary results of modelling causal arguments based on the legal decisions with particular focus on physical causation. We introduce a semi-formal framework for reasoning with causation that uses strict and defeasible rules for modelling factual causation arguments in legal cases. We further discuss the complex relation between formal, common sense, norm and policy based considerations of causation in legal decision making with particular focus on their role in comparing alternative causal explanations.

This paper presents the features of a moderately simple legal expert system devoted to solving the most frequent legal problems of disabled persons in Poland. The authors focused on the structure of legal expert system and methodology used for the sake of its development. The succession law of Poland has been selected in the paper as the illustrative domain, because the modelling of the succession procedures delivers sufficient material to reveal the most important issues concerning project of the legal expert system.

This work reports on the ongoing development of a multilingual pipeline for the summarization of European case law. We apply the TextRank algorithm on concepts of the EuroVoc thesaurus in order to extract summarizing keywords and sentences. In a first case study, we demonstrate the feasibility and usefulness of the presented approach for five different languages and 18 document sources.

legivoc is the first Internet-based platform dedicated to the diffusion, edition and alignment of legal vocabularies across countries. Funded in part by the European Commission and administered by the French Ministry of Justice, legivoc offers a seamless path for governments to disseminate their legal foundations and specify semantic bridges between them. In this system description paper, we outline the general principles behind the legivoc framework and provide some ideas about its implementation, with a particular focus on the state-of-the-art tools it includes to help crowdsource the alignment of legal corpora together.

Ambient Assisted Living provides solutions to the increasing cognitive problems that affect the elderly population. To provide all features possible, Ambient Assisted Living projects require access to personal and private information of their users. Currently, the legal issues arisen in Ambient Assisted Living are a hot topic in the European Union, especially aspects regarding unsupervised data processing and cross-sharing of personal information. In this paper it is presented the iGenda project, which is a Cognitive Assistant inserted in the Ambient Assisted Living area which aims to build safe environments that adapt themselves to one’s individual needs. However, one of the issues is the protection of the data flowing within the system and the protection of user’s fundamental rights. It is also presented the principles and legal guarantees of data protection and transmission, and legal aspects are explained, embracing appropriate solutions to technological features that may be a threat.

This paper discusses the problem of lack of clear licensing and transparency of usage terms and conditions for research metadata. Making research data connected, discoverable and reusable are the key enablers of the new data revolution in research. We discuss how the lack of transparency hinders discovery of research data and make it disconnected from the publication and other trusted research outcomes. In addition, we discuss the application of Creative Commons licenses for research metadata, and provide some examples of the applicability of this approach to internationally known data infrastructures.

The result of the EU is a complex, multilingual, multicultural and yet united environment, requiring solid integration policies and actions targeted at simplifying cross-language and cross-cultural knowledge access. The legal domain is a typical case in which both the linguistic and the conceptual aspects mutually interweave into a knowledge barrier that is hard to break. In the context of the ISA2 funded project “Public Multilingual Knowledge Infrastructure” (PMKI) we are addressing Semantic Interoperability at both the conceptual and lexical level, by developing a set of coordinated instruments for advanced lexicalization of RDF resources (be them ontologies, thesauri and datasets in general) and for alignment of their content. In this paper, we describe the objectives of the project and the concrete actions, specifically in the legal domain, that will create a platform for multilingual cross-jurisdiction accessibility to legal content in the EU.

The paper introduces a Multilingual Access Module. This module translates the user’s legislation query from its source language into the target language, and retrieves the detected texts that match the query. The service is demonstrated in its potential for two languages – English and Bulgarian, in both directions (English-to-Bulgarian and Bulgarian-to-English). The module consists of two submodules: Ontology-based and Statistical Machine Translation. Since both proposed submodules have some drawbacks, they are used in an integrated architecture, thus profiting from each other.

Legal texts express conditions in natural language describing what is permitted, forbidden or mandatory in the context they regulate. Despite the numerous approaches tackling the problem of moving from a natural language legal text to the respective set of machine-readable conditions, results are still unsatisfiable and it remains a major open challenge. In this paper, we propose a preliminary approach which combines different Natural Language Processing techniques towards the extraction of rules from legal documents. More precisely, we combine the linguistic information provided by WordNet together with a syntax-based extraction of rules from legal texts, and a logic-based extraction of dependencies between chunks of such texts. Such a combined approach leads to a powerful solution towards the extraction of machine-readable rules from legal documents. We evaluate the proposed approach over the Australian “Telecommunications consumer protections code”.

The earthquake struck in Emilia-Romagna Region on 2012 and it created a disaster area with 33 municipalities involved and extended over 3,173 square meters across the region. The Commissioner for Emergencies issued 350 ordinances deliberated over a three-year period (2012–2015), so as to support the rebuilding, aid for the population, organization of the territory. The main goal of this paper is to present the outcome of a research that investigated the corpus of legislative ordinances in the first 18 months in order to discover if they were an effective legislative instrument in emergency settings. Analyzing the legal citations and the correspondent references we have discovered some dysfunctional behaviour in the lawmaking system, too much concentrated on some topics. We have detected weaknesses in normative area that could orient a more coordinated legislative action at the national level. The final findings help the lawmaker act better in future disasters, extract information concerning the number and the types of modifications produced, and support the debate on a national law on emergency in the wake of natural disasters.

Law is an explicit system of rules to govern the behaviour of people. Legal practitioners must learn to apply legal knowledge to the facts at hand. The United States Multistate Bar Exam (MBE) is a professional test of legal knowledge, where passing indicates that the examinee understands how to apply the law. This paper describes an initial attempt to model and implement the automatic application of legal knowledge using a rule-based approach. An NLP tool extracts information (e.g. named entities and syntactic triples) to instantiate an ontology relative to concepts and relations; ontological elements are associated with legal rules written in SWRL to draw inferences to an exam question. The preliminary results on a small sample are promising. However, the main development is the methodology and identification of key issues for future analysis.

The use of pragmatics in applying the law is hard to deal with for a legal knowledge engineer who needs to model it in a precise KR for (semi-)automated legal reasoning systems. The negative aspects of pragmatics is due to the difficulty involved in separating their concerns. When representing a legal norm for (semi-)automated reasoning, an important step/aspect is the annotation of legal sections under consideration. Annotation in the context of this paper refers to identification, segregation and thereafter representation of the content and its associated context. In this paper we present an approach and provide a proof-of-concept implementation for automatizing the process of identifying the most relevant judgment pertaining to a legal section and further transforming them into a formal representation format. The annotated legal section and its related judgments can then be mapped into a decision model for further down the line processing.

This paper describes a task of semantic labeling of document segments. The idea exploits ontology in providing a fine-grained conceptual document annotation. We describe a way of dividing a document into its constituent semantically-coherent blocks. These blocks are then used to perform conceptual tagging for efficient passage information retrieval. The proposed task interfaces other application areas such as intra-mapping of ontologies, text summarization and information extraction. The system has been evaluated on a task of conceptual tagging of documents and achieved a promising result.

Instead of custom-building a new ontology from scratch, knowledge resources can be elicited, reused and engineered to develop legal ontologies with the goal of promoting the application of good practices and speeding up the ontology development process. This paper focuses on the specificities of non-ontological resources in the legal domain, and provides some guidelines of how these can be reused and engineered to enable heterogeneous resources integration within a legal ontology. The paper presents some examples of these processes using a case-study in the consumer law domain.

In the continental law system, more attention is paid to judicial interpretation to judge legal facts or actions than judicial precedents. Therefore, in the continental legal system, it is appropriate to express the law itself with knowledge such as legal ontologies or logical rules. To construct legal ontologies and rule-based methods, legal analysis by collaboration between legal experts and knowledge engineers should be preceded. This paper proposes a general model for designing criminal law ontologies and rules. First, we introduce the super-domain ontology that contains the common characteristics of criminal law. Then, we explain the rule design method of criminal law and present the application of the anti-graft act in Korea as an example.

The Italian anti-corruption Act (law n. 190/2012) requires all public administrations to spread procurement information as open data. Each body is obliged to yearly release standardized XML files, on its public website, containing data that describes all issued public contracts. Though this information is currently available on a machine-readable format, the data is fragmented and published in different files on different websites, without a unified and human-readable view of the information. The ContrattiPubblici.org project aims at developing a semantic knowledge graph based on linked data principles in order to overcome the fragmentation of existent datasets, to allow easy analysis, and to enable the reuse of information. The objectives are to increase public awareness about public spending, to improve transparency on the public procurement chain, and to help companies to retrieve useful knowledge for their business activities.

The Ontology modularization is an essential field in the ontology engineering domain helping to reduce the complexity and the difficulties of building, reusing, managing and reasoning on domain ontologies either by applying partitioning or composition approaches. This paper carries out a survey on ontology modularization and presents a modular approach to build criminal modular domain ontology (CriMOnto) for modelling the legal norms of the Lebanese criminal system. CriMOnto, which will be used later for a legal reasoning system, is composed of four independent modules. The modules will be combined together to compose the whole ontology.

Ontolex-lemon is the de facto standard to represent lexica relative to ontologies and it can be used to encode term banks as RDF. A multi-lingual, multi-jurisdictional term bank of copyright-related concepts has been published as linked data based on the ontolex-lemon model. The terminology links information from WIPO (concepts and definitions), IATE (multilingual terms, usage notes) and other sources as Creative Commons (multilingual definitions) or DBpedia (general concepts). The terms have been hierarchically arranged, spanning multiple languages and targeting different jurisdictions. The term bank has been published as a TBX dump file and is publicly accessible as linked data. The term bank has been used to annotate common licenses in the RDFLicense dataset.

Triple-A is an abstract argumentation model, distinguishing the global argumentation of judges from the local argumentation of accused, prosecutors, witnesses, lawyers, and experts. In Triple-A, agents have partial knowledge of the arguments and attacks of other agents, and they decide autonomously whether to accept or reject their own arguments, and whether to bring their arguments forward in court. The arguments accepted by the judge are based on a game-theoretic equilibrium among the argumentation of the other agents. The Triple-A theory can be used to distinguish various direct and indirect ways in which the arguments of an agent can be used against his or her other arguments.

This study aims to analyze and evaluate the natural language arguments present in legal documents. The research is divided into three modules or stages: an Argument Element Identifier Module identifying argumentative and non-argumentative sentences in legal texts; an Argument Builder Module handling clustering of argument’s components; and an Argument Structurer Module distinguishing argument’s components (premises and conclusion). The corpus selected for this research was the set of Case-Laws issued by the European Court of Human Rights (ECHR) annotated by Mochales-Palau and Moens [8]. The preliminary results of the Argument Element Identifier Module are presented, including its main features. The performance of two machine learning algorithms (Support Vector Machine Algorithm and Random Forest Algorithm) is also measured.

This paper highlights the benefit of semantic information retrieval in legal networks. User queries get more complex when they combine constraints on semantic content and intertextual links between documents. Comparing two methods of search in legal collection networks, we present new functionalities of search and browsing. Relying on a structured representation of the collection graph, the first approach allows for approximate answers and knowledge discovery. The second one supports richer semantics and scalability but offers fewer search functionalities. We indicate how those approaches could be combined to get the best of both.

Administrative adjudications are the most common form of legal decisions in many countries, so improving the efficiency, accuracy, and consistency of administrative processes could significantly benefit agencies and citizens alike. We explore the hypothesis that predictive models induced from previous administrative decisions can improve subsequent decision-making processes. This paper describes three datasets for exploring this hypothesis: motion-rulings, Board of Veterans Appeals (BVA) decisions; and World Intellectual Property Organization (WIPO) domain name dispute decisions. Three different approaches for prediction in these domains were tested: maximum entropy over token n-grams; SVM over token n-grams; and a Hierarchical Attention Network (HAN) applied to the full text. Each approach was capable of predicting outcomes, with the simpler WIPO cases appearing to be much more predictable than BVA or motion-ruling cases. We explore several approaches to using predictive models to identify salient phrases in the predictive texts (i.e., motion or contentions and factual background) and propose a design for incorporating this information into a decision-support tool.

Many researchers have worked on formalizing legal reasoning and the representation of law. Particularly in the last decade progress has been made in creating formal models of argumentation. We aim to develop an approach that is not only formally correct, but also can be used and understood by common legal practitioners and IT-staff members. The approach should provide an instrument that can be used to inform legal experts on relevant issues when finding a solution to a case at hand. In this paper, we present a real-time case that is in discussion within the Dutch Tax Administration, the Ministry of Finance, in court as well as in Parliament. By making a structured interpretation of sources of norms relevant for the case, and argument schemes to make differences of opinion explicit, we aim to demonstrate the FLINT method for the interpretation of norms, and support a process of redesign of administrative procedures.

We present a novel activating approach to Argument Theory Change (ATC) for the study of the dynamics of the judicial process. ATC applies belief change concepts to dialectical argumentation for altering trees upon which the semantics for reasoning are defined. The activating approach to ATC considers the incorporation of arguments to define a revision operator for studying how to provoke change to the semantics’ outcome. Our objective is to contribute to the discussion of how to deal with circumstances of the judicial process like hypothetical reasoning for conducting investigations of a legal case, and for handling the dynamics of the judicial process. We finally observe the behavior of our proposal upon the sentences of two different real criminal procedures.

Mining arguments from text has recently become a hot topic in Artificial Intelligence. The legal domain offers an ideal scenario to apply novel techniques coming from machine learning and natural language processing, addressing this challenging task. Following recent approaches to argumentation mining in juridical documents, this paper presents two distinct contributions. The first one is a novel annotated corpus for argumentation mining in the legal domain, together with a set of annotation guidelines. The second one is the empirical evaluation of a recent machine learning method for claim detection in judgments. The method, which is based on Tree Kernels, has been applied to context-independent claim detection in other genres such as Wikipedia articles and essays. Here we show that this method also provides a useful instrument in the legal domain, especially when used in combination with domain-specific information.

There is a consensus that trust in one’ opponent plays a significant role in promoting parties to engage in the conflict management process. Trust is an important yet complex and little-understood relation among parties in conflict. In general, trust can be seen as a measure of confidence that an entity or entities will behave expectedly. Without trust, the instruments to prevent or manage the conflict, such as negotiation, are handicapped and cannot reach their full potential for promoting an end to or a mitigation of a conflict. Hence, our motivation to examine trust is three-fold. First, the present study aims to address and expand on this line of research by investigating the possibility of measuring trust based on quantifiable behavior. To do so, we provide a brief review of the existing definitions of trust and define trust in the context of a negotiation scenario. Further, we propose a formal definition so that the analysis of trust in this kind of scenarios can be developed. Thus, it is suggested the use of Ambient Intelligence techniques that use a trust data model to collect and evaluate relevant information based on the assumption that observable trust between two entities (parties) results in certain typical behaviors. Third, this work aims to study the particular connection between relational aspects of trust and parties’ conflict styles based on two dimensions: cooperativeness and assertiveness. The main contribution of this work is the identification of situations in which trust relationships influences the negotiation performance. To do so, an experiment was set-up in which we tried to streamline all the relevant aspects of the interaction between the parties and its environment that occur in a sensory rich environment, to measure trust. To simulate a conflict situation, a web-based game was developed. It was designed to enable test participants to engage in a conflict experience induced by the presence of Ambient Intelligence systems. Several tests were performed. We then engaged in rigorous assessment, post- processing and analysis of results. We validated the results comparing them with trust measures obtained through the use of a questionnaire (carefully adapted) from social networks.

Legal Informatics has recently witnessed a growing interest towards the insights offered by the intersection among Network Analysis (NA), visualization techniques and legal science research questions. Also thanks to several seminal works, the field is ready to tackle new challenges at a theoretical and application level. The first is to bring the network approach into “genuinely legal” research questions. The second is to create tools allowing legal scholars without technical skills to exploit NA with two goals: (i) make experiments with NA and push new ideas both in legal and NA science; (ii) use NA and visualization in their daily activities (e.g., legal analysis and information retrieval). Against this backdrop, a truly interdisciplinary approach deeply involving legal experts/scholars is needed. The paper presents an ongoing research project - EUCaseNet - dealing with these challenges and aiming to explore the potentialities of NA in supporting the study of EU case law.

In a cross-border dimension considering the specific collaboration among European Union Member States related to criminal investigations and criminal trials, it becomes crucial to have a common and shared understanding of what Electronic Evidence is and how it should be treated in the EU context and in the EU MS. In this context the EVIDENCE project developed a tailor-made categorization of relevant concepts that allow to rely on a common and shared knowledge in this domain. The categorization provides a starting analysis for the exchange of Electronic Evidence and data between judicial actors and LEAs, with a specific focus on issues of the criminal field and criminal procedures. This semantic Structure might represent a good starting point for the alignment of electronic evidence concepts all over Europe in across border dimension. This categorisation is significant as it is one of the few initiatives to identify and classify relevant concepts in a domain, which currently lacks of clear boundaries and touches upon different disciplines.