Search

This article describes the creation of a lightweight ontology of European Union (EU) criminal procedural rights in judicial cooperation. The ontology is intended to help legal practitioners understand the precise contextual meaning of terms as well as helping to inform the creation of a rule ontology of criminal procedural rights in judicial cooperation. In particular, we started from the problem that directives sometimes do not contain articles dedicated to definitions. This issue provided us with an opportunity to explore a phenomenon typically neglected in the construction of domain-specific legal ontologies. Whether classical definitions are present or absent, laws and legal sources in general are typically peppered with a number of hidden definitions (in the sense that they are not clearly marked out as such) as well as incomplete definitions, which may nevertheless help legal practitioners (and legal reasoning systems) to reason on the basis of analogy or teleology. In this article we describe the theoretical basis for building an analogical lightweight ontology in the framework of an EU project called CrossJustice. We present our methodology for collecting the data, extracting the data fields and creating the ontology with WebProtégé, followed by our conclusions and ideas for future work.

This article seeks to address the problem of the ‘resource consumption bottleneck’ of creating legal semantic technologies manually. It describes a semantic role labeling based information extraction system to extract definitions and norms from legislation and represent them as structured norms in legal ontologies. The output is intended to help make laws more accessible, understandable, and searchable in a legal document management system.

Nowadays, many companies are offering chatbots and voicebots to their customers. Despite much recent success in natural language processing and dialogue research, the communication between a human and a machine is still in its infancy. In this context, dialogue personalization could be a key to bridge some of the gap, making sense of users’ experiences, needs, interests and mental models when engaged in a conversation. On this line, we propose to automatically learn user’s features directly from the dialogue with the chatbot, in order to enable the adaptation of the response accordingly and thus improve the interaction with the user. In this paper, we focus on the user’s domain expertise and, assuming that expertise affects linguistic features of the language, we propose a vocabulary-centered model joint with a Deep Learning method for the automatic classification of the users expertise at word- and message-level. An experimentation over 5000 real messages taken from a telco commercial chatbot carried to high accuracy scores, demonstrating the feasibility of the proposed task and paving the way for novel user-aware applications.

To date, the effort made by existing vocabularies to provide a shared representation of the data protection domain is not fully exploited. Different natural language processing (NLP) techniques have been applied to the text of privacy policies without, however, taking advantage of existing vocabularies to provide those documents with a shared semantic superstructure. In this paper we show how a recently released domain-specific vocabulary, i.e. the Data Privacy Vocabulary (DPV), can be used to discover, in privacy policies, the information that is relevant with respect to the concepts modelled in the vocabulary itself. We also provide a machine-readable representation of this information to bridge the unstructured textual information to the formal taxonomy modelled in it. This is the first approach to the automatic processing of privacy policies that relies on the DPV, fuelling further investigation on the applicability of existing semantic resources to promote the reuse of information and the interoperability between systems in the data protection domain.

The automated identification of national implementations (NIMs) of European directives by text similarity techniques has shown promising preliminary results. Previous works have proposed and utilized unsupervised lexical and semantic similarity techniques based on vector space models, latent semantic analysis and topic models. However, these techniques were evaluated on a small multilingual corpus of directives and NIMs. In this paper, we utilize word and paragraph embedding models learned by shallow neural networks from a multilingual legal corpus of European directives and national legislation (from Ireland, Luxembourg and Italy) to develop unsupervised semantic similarity systems to identify transpositions. We evaluate these models and compare their results with the previous unsupervised methods on a multilingual test corpus of 43 Directives and their corresponding NIMs. We also develop supervised machine learning models to identify transpositions and compare their performance with different feature sets.

Natural language processing techniques often aim at automatically extracting semantics from texts. However, they usually need some available semantic knowledge contained in dictionaries and resources such as WordNet, Wikipedia, and FrameNet. In this respect, there is a large literature about the creation of novel semantic resources as well as attempts to integrate existing ones. In this context, we here focus on common-sense knowledge, which shows to have interesting characteristics as well as challenging issues such as ambiguity, vagueness, and inconsistency. In this paper, we make use of a large-scale and crowdsourced common-sense knowledge base, i.e., ConceptNet, to qualitatively evaluate its role in the perception of semantic association among words. We then propose an unsupervised method to disambiguate and integrate ConceptNet instances into WordNet, demonstrating how the enriched resource improves the recognition of semantic association. Finally, we describe a novel approach to label semantically associated words by exploiting the functional and behavioral information usually contained in common sense, demonstrating how this enhances the explanation (and the use) of relatedness and similarity with non-numeric information.

Because of the recent entry into force of the General Data Protection Regulation (GDPR), a growing of documents issued by the European Union institutions and authorities often mention and discuss various use cases to be handled to comply with GDPR principles. This contribution addresses the problem of extracting recurrent use cases from legal documents belonging to the data protection domain by exploiting existing Ontology Design Patterns (ODPs). An analysis of ODPs that could be looked for inside data protection related documents is provided. Moreover, a first insight on how Natural Language Processing techniques could be exploited to identify recurrent ODPs from legal texts is presented. Thus, the proposed approach aims to identify standard use cases in the data protection field at EU level to promote the reuse of existing formalisations of knowledge.

International audience; Ontologies represent the standard way to model the knowledge about specific domains. This holds also for the legal domain where several ontologies have been put forward to model specific kinds of legal knowledge. Both for standard users and for law scholars, it is often difficult to have an overall view on the existing alternatives, their main features and their interlinking with the other ontologies. To answer this need, in this paper, we address an analysis of the state-of-the-art in legal ontologies and we characterise them along with some distinctive features. This paper aims to guide generic users and law experts in selecting the legal ontology that better fits their needs and in understanding its specificity so that proper extensions to the selected model could be investigated.

The continuous evolution of research has led to an exponential growth of the scientific literature. This engenders difficulties for researchers to entirely capture the most salient efforts related to their own research. In this paper, we propose a novel knowledge model for unveiling meaningful and labeled relations among articles based on both topics and latent citation dependencies. An experimentation on the whole literature in the Computer Science field allowed us to validate our approach by bridging the gap between few lines of textual content (e.g., an abstract) to the most relevant papers to be included in the bibliography.

In recent years, new digital technologies are being used to support the navigation and the analysis of scientific publications, justified by the increasing number of articles published every year. For this reason, experts make use of on-line systems to browse thousands of articles in search of relevant information. In this paper, we present a new method that automatically assigns meanings to references on the basis of the citation text through a Natural Language Processing pipeline and a slightly-supervised clustering process. The resulting network of semantically-linked articles allows an informed exploration of the research panorama through semantic paths. The proposed approach has been validated using the ACL Anthology Dataset containing several thousands of papers related to the Computational Linguistics field. A manual evaluation on the extracted citation meanings carried to very high levels of accuracy. Finally, a freely-available web-based application has been developed and published on-line.

It is often hard to apply the newest advances in research to real life scenarios. They usually require the resolution of some specific task applied to a restricted domain, all the while providing small amounts of data to begin with. In this study we apply one of the newest innovations in Deep Learning to a task of text classification. We created a question answering system in Italian that provides information about a specific subject, e-invoicing and digital billing. Italy recently introduced a new legislation about e-invoicing and people have some legit doubts, therefore a large share of professionals could benefit from this tool.

Concepts and relations within existing ontologies usually represent limited subjective and application-oriented views of a domain of interest. However, reusing resources and fine-grained conceptualizations is often challenging and requires significant manual efforts of adaptation to fit with unprecedented usages. In this paper, we present a system that makes use of recent open information extraction technologies to unravel and explore corpus-centered unknown relations in the legal domain.keywordsopen information extractionnatural language processingontologieslegal conceptslegal textiate.

Natural language development and learning begins with a few terms, often related to concepts associated to perception, in general neither too general nor too specific, gradually specializing and extending to abstract concepts. A linguistic middle-level can be considered as a lexical base on which a language oriented towards more complex communications develops in adulthood. Learning can extend from generic to specific terms (usually a child learns first the term "dog", then the term "dachshund"), or from specific to generic ("dog" → "mammal"). Thus, a middle-level includes all those terms that are halfway, in the taxonomic sense, between generic and specific concepts. In this paper we propose a computational approach to identify a middle-level relying on Wikipedia as input corpus and WordNet as lexical-semantic resource. An experimentation based on graded readings shows favourable results. The impact of this work could touch different fields, e.g., the automatic evaluation of the complexity of a text by computational approaches, rather than psychological and linguistic studies related to the language use.

This paper presents a model for school remedial, focusing on improving the digital materials sharing process for the diversification of tutoring strategies. The model involves the characterization of materials for automatic assessment shared within a community of tutors. The characterization expects materials to be linked with natural language descriptors explicating their intended instructional objectives. The possibility of implementing a recommendation system on the basis of natural language processing techniques is discussed taking in consideration an analysis of the application of the model within a local-scale project. Clustering techniques searching for materials that have the same educational purposes but involve the activation of different cognitive processes are proposed, in order to continuously extend the database of shared materials in favour of the effectiveness of ongoing tutoring actions. The results collected from questionnaires submitted to students, tutors, and teachers involved in the project are shown, and clustering data are discussed highlighting the feasibility of the application of the model.

The emergence of Technology Enhanced Learning environments has led to the continual growth of the availability of digital educational resources. In this paper, the potential of enabling their reuse into student-centric services – such as recommender systems or adaptive tutoring tools – is discussed through the proposal and comparison of procedures for automatically detecting the mutual relatedness among learning objects. Since the choice of the similarity measure is fundamental for clustering digital materials, this paper addresses the investigation on two distinct approaches: the content-based semantic similarity, compared to the closeness measure on natural language descriptions of metadata – namely prerequisites and educational objectives. The analysis is conducted on a collection of mathematical problems, equipped with metadata which facilitate their retrieval in Virtual Learning Environments, created by Secondary School teachers with the support of University experts. Natural Language Processing techniques are exploited for extracting relevant information from the metadata, while the developments in the emergent field of Mathematical Language Processing are proposed for the treatment of mathematical expressions included in the resources. The distinct similarity measures presented are examined considering the compared results, and their correlation is evaluated. This study is intended to be the first step towards the definition of a model for structuring shared materials available in disciplinary repositories of virtual communities. This model will be used for implementing a system for the delivery of learning objects trajectories on a digital map automatically generated. The system’s efficacy will be tested through its integration to a Learning Management System hosting secondary school classrooms’ courses. The research is part of a PhD in Pure and Applied Mathematics in apprenticeship, conducted in partnership with leading providers of software based on Computer Algebra System engine.

This paper presents a description about our adopted approach for the information retrieval and textual entailment tasks of the COLIEE 2017 competition. We address the information retrieval task by implementing a partial string matching and a topic clustering method. For the textual entailment task, we propose a Long Short-Term Memory (LSTM) - Convolutional Neural Network (CNN) model which utilizes word embeddings trained on the Google News vectors. We evaluated our approach for both tasks on the COLIEE 2017 dataset. The results demonstrate that the topic clustering method outperformed the partial string matching method in the information retrieval task. The performance of LSTM-CNN model was competitive with other textual entailment systems.

In this paper we present the BO-ECLI Parser, an open framework for the extraction of legal references from case-law issued by judicial authorities of European member States. The problem of automatic legal links extraction from texts is tackled for multiple languages and jurisdictions by providing a common stack which is customizable through pluggable extensions in order to cover the linguistic diversity and specific peculiarities of national legal citation practices. The aim is to increase the availability in the public domain of machine readable references meta-data for case-law by sharing common services, a guided methodology and efficient solutions to recurrent problems in legal references extraction, that reduce the effort needed by national data providers to develop their own extraction solution.

This paper presents a model for structuring shared resources, proposed to activate learners' formative and proactive assessment processes by enhancing the sharing-workflow among instructors. The model is based on the features of a Virtual Learning Community (VLC), and provides for the development and reuse of ontologies. The adoption of the model for the implementation of an adaptive system for the dispatching of materials is discussed on the basis of the results from clustering analyses. Semantic-similarity measures are compared by their application for clustering a collection of mathematical questions for automatic assessment, created by instructors within a national-wide VLC for Secondary Schools.

Legal ontologies aim to provide a structured representation of legal concepts and their interconnections. These ontologies are then exploited to support information extraction and question answering in the legal domain. In addition, given the increasing importance of the Web of Data in public administration and in companies, being able to provide machine-readable legal information is becoming a valuable and desired contribution. The problem is that these ontologies are not always very transparent to end users, in particular if they lack legal knowledge. In this context, we present the InvestigatiOnt tool which aims to ease the interaction of end users with legal ontologies in order to spread the use of machine-processable legal information as well as its understanding.

This paper presents an ontological model for defining competency paths in STEM education, designed for the implementation of an adaptive system integrated in virtual communities. The model is applied for clustering materials for automatic assessment and the results are discussed.

Nowadays, there is a huge amount of textual data coming from on-line social communities like Twitter or encyclopedic data provided by Wikipedia and similar platforms. This Big Data Era created novel challenges to be faced in order to make sense of large data storages as well as to efficiently find specific information within them. In a more domain-specific scenario like the management of legal documents, the extraction of semantic knowledge can support domain engineers to find relevant information in more rapid ways, and to provide assistance within the process of constructing application-based legal ontologies. In this work, we face the problem of automatically extracting structured knowledge to improve semantic search and ontology creation on textual databases. To achieve this goal, we propose an approach that first relies on well-known Natural Language Processing techniques like Part-Of-Speech tagging and Syntactic Parsing. Then, we transform these information into generalized features that aim at capturing the surrounding linguistic variability of the target semantic units. These new featured data are finally fed into a Support Vector Machine classifier that computes a model to automate the semantic annotation. We first tested our technique on the problem of automatically extracting semantic entities and involved objects within legal texts. Then, we focus on the identification of hypernym relations and definitional sentences, demonstrating the validity of the approach on different tasks and domains.

Nowadays, Sentiment Analysis (SA) is receiving huge attention because of the wide range of its direct applications like analyses of products, customer profiles, political trends, and so forth. Still, the availability of big amounts of data coming from the World Wide Web makes easier the study of both new techniques and evaluation methods. Current literature mainly focuses on two approaches which rely on sentiment lexicons (i.e., lists of words associated to scores of sentiment polarity) or on Natural Language Processing techniques (NLP). In this paper, on one hand, we introduce and evaluate a novel algorithm for SA that relies on a simple set of propagation rules applied at syntactic level within a dependency parse tree. On the other hand, we propose a context-based model where the users' sentiments (or opinions) are tuned according to some context of analysis. Finally, we present the system called SentiVis which implements these ideas through an orthogonal approach to SA that directly leans on Data Visualization. Extracted sentiments, with respect to some query of analysis, are ordered and represented graphically in a 2-dimensional space, conveying information about their strength and variability. This way, we avoid cumbersome rankings of objects and associated opinions by directly mapping such information on the screen. The user is then able to interact with the visualized data in order to discover interesting facts as well as removing false positive (or negative) opinions deriving by the used algorithm. We then evaluate the efficacy of the proposed system through several case studies.

This paper presents a unifying text similarity measure (USM) for automated identification of national implementations of European Union (EU) directives. The proposed model retrieves the transposed provisions of national law at a fine-grained level for each article of the directive. USM incorporates methods for matching common words, common sequences of words and approximate string matching. It was used for identifying transpositions on a multilingual corpus of four directives and their corresponding national implementing measures (NIMs) in three different languages : English, French and Italian. We further utilized a corpus of four additional directives and their corresponding NIMs in English language for a thorough test of the USM approach. We evaluated the model by comparing our results with a gold standard consisting of official correlation tables (where available) or correspondences manually identified by domain experts. Our results indicate that USM was able to identify transpositions with average F-score values of 0.808, 0.736 and 0.708 for French, Italian and English Directive-NIM pairs respectively in the multilingual corpus. A comparison with state-of-the-art methods for text similarity illustrates that USM achieves a higher F-score and recall across both the corpora.

This paper describes a system which uses entity and topic coherence for improved Text Segmentation (TS) accuracy. First, the Linear Dirichlet Allocation (LDA) algorithm was used to obtain topics for sentences in the document. We then performed entity mapping across a window in order to discover the transition of entities within sentences. We used the information obtained to support our LDA-based boundary detection for proper boundary adjustment. We report the significance of the entity coherence approach as well as the superiority of our algorithm over existing work.

In the last decades, many measures and metrics have been proposed with the goal of automatically providing quantitative rather than qualitative indications over researchers' academic productions. However, when evaluating a researcher, most of the commonly-applied measures do not consider one of the key aspect of every research work: the collaborations among researchers and, more specifically, the impact that each co-author has on the scientific production of another. In fact, in an evaluation process, some co-authored works can unconditionally favor researchers working in competitive research environments surrounded by experts able to lead high-quality research projects, where state-of-the-art measures usually fail in trying to distinguish co-authors from their pure publication history. In the light of this, instead of focusing on a pure quantitative/qualitative evaluation of curricula, we propose a novel temporal model for formalizing and estimating the dependence of a researcher on individual collaborations, over time, in surrounding communities. We then implemented and evaluated our model with a set of experiments on real case scenarios and through an extensive user study.

In this paper we present the system KnowNow, a tool whose aim is to let the users navigate into text corpora through dynamic semantic information networks, created in real-time according to delimited time ranges. In educational scenarios, students are often asked to write short essays on different topics linked by temporal information. This usually involves a combination of several aspects to be evaluated, such as knowledge, imagination, structure and presentation. In the light of this, the introduction of Natural Language Understanding techniques together with cross-topic navigation and visualization tools and considerably help students to retrieve, link, and create well-structured and original contributions, as we demonstrate by using KnowNow.

In this paper we address the problem of automatically constructing structured knowledge from plain texts. In particular, we present a supervised learning technique to first identify definitions in text data, while then finding hypernym relations within them making use of extracted syntactic structures. Instead of using pattern matching methods that rely on lexico-syntactic patterns, we propose a method which only uses syntactic dependencies between terms extracted with a syntactic parser. Our assumption is that syntax is more robust than patterns when coping with the length and the complexity of the texts. Then, we transform the syntactic contexts of each noun in a coarse-grained textual representation, that is later fed into hyponym/hypernym-centered Support Vector Machine classifiers. The results on an annotated dataset of definitional sentences demonstrate the validity of our approach overtaking the current state of the art.

In this paper, we describe how NLP can semi-automate the construction and analysis of knowledge in Eunomos, a legal knowledge management service which enables users to view legislation from various sources and ?nd the right de?nitions and explanations of legal concepts in a given context. NLP can semi-automate some routine tasks currently performed by knowledge engineers, such as classifying norms, or linking key terms within legislation to ontological concepts. This helps overcome the resource bottleneck problem of creating specialist knowledge management systems. While accuracy is of the utmost importance in the legal domain, and the information should be veri?ed by domain experts as a matter of course, a semi-automated approach can result in considerable ef?ciency gains.

The Radial Coordinate Visualization (Radviz) technique has been widely used to effectively evaluate the existence of patterns in highly dimensional data sets A crucial aspect of this technique lies in the arrangement of the dimensions, which determines the quality of the posterior visualization Dimension arrangement (DA) has been shown to be an NP-problem and different heuristics have been proposed to solve it using optimization techniques However, very little work has focused on understanding the relation between the arrangement of the dimensions and the quality of the visualization In this paper we first present two variations of the DA problem: (1) a Radviz independent approach and (2) a Radviz dependent approach We then describe the use of the Davies-Bouldin index to automatically evaluate the quality of a visualization i.e., its visual usefulness Our empirical evaluation is extensive and uses both real and synthetic data sets in order to evaluate our proposed methods and to fully understand the impact that parameters such as number of samples, dimensions, or cluster separability have in the relation between the optimization algorithm and the visualization tool.

We present a demonstration of ClusTR, a highly interactive system for exploring relationships between different clusterings of a dataset and for viewing the evolution in time of topics (e.g., tags associated with objects in the dataset) within and across such clusters. In particular, ClusTR allows exploration of generic multi-dimensional, text labeled and time sensitive data.

We present the tagFlake system, which supports semantically informed navigation within a tag cloud. tagFlake relies on TMine for organizing tags extracted from textual content in hierarchical organizations, suitable for navigation, visualization, classification, and tracking. TMine extracts the most significant tag/terms from text documents and maps them onto a hierarchy in such a way that descendant terms are contextually dependent on their ancestors within the given corpus of documents. This provides tagFlake with a mechanism for enabling navigation within the tag space and for classification of the text documents based on the contextual structure captured by the created hierarchy. tagFlake is language neutral, since it does not rely on any natural language processing technique and is unsupervised.

This paper describes the Eunomos software, an advanced legal document and knowledge management system, based on legislative XML and ontologies. We describe the challenges of legal research in an increasingly complex, multi-level and multi-lingual world and how the Eunomos software helps users cut through the information overload to get the legal information they need in an organized and structured way and keep track of the state of the relevant law on any given topic. Using NLP tools to semi-automate the lower-skill tasks makes this ambitious project a realistic commercial prospect as it helps keep costs down while at the same time allowing greater coverage. We describe the core system from workflow and technical perspectives, and discuss applications of the system for various user groups.

This paper presents a deep learning based dialogue system which has been trained to answer user queries posed as questions during a conversation. The proposed system, though generative, takes advantage of domain specific knowledge for generating valid answers. The evaluation analysis shows that the proposed system obtained a promising result.