Last 7 days
Bookmark and Share    
Full Text
Peer Reviewed
See detailThe DAta Protection REgulation COmpliance Model
Bartolini, Cesare UL; Lenzini, Gabriele UL; Robaldo, Livio UL

in IEEE Security and Privacy (2019), 17(6), 37-45

Understanding whether certain technical measures comply with the General Data Protection Regulation’s (GDPR’s) principles is complex legal work. This article describes a model of the GDPR that allows for ... [more ▼]

Understanding whether certain technical measures comply with the General Data Protection Regulation’s (GDPR’s) principles is complex legal work. This article describes a model of the GDPR that allows for semiautomatic processing of legal text and the leveraging of state-of-the-art legal informatics approaches, which are useful for legal reasoning, software design, information retrieval, or compliance checking. [less ▲]

Detailed reference viewed: 98 (7 UL)
Full Text
Peer Reviewed
See detailTemporal 3D Human Pose Estimation for Action Recognition from Arbitrary Viewpoints
Adel Musallam, Mohamed; Baptista, Renato UL; Al Ismaeil, Kassem UL et al

in 6th Annual Conf. on Computational Science & Computational Intelligence, Las Vegas 5-7 December 2019 (2019, December)

This work presents a new view-invariant action recognition system that is able to classify human actions by using a single RGB camera, including challenging camera viewpoints. Understanding actions from ... [more ▼]

This work presents a new view-invariant action recognition system that is able to classify human actions by using a single RGB camera, including challenging camera viewpoints. Understanding actions from different viewpoints remains an extremely challenging problem, due to depth ambiguities, occlusion, and a large variety of appearances and scenes. Moreover, using only the information from the 2D perspective gives different interpretations for the same action seen from different viewpoints. Our system operates in two subsequent stages. The first stage estimates the 2D human pose using a convolution neural network. In the next stage, the 2D human poses are lifted to 3D human poses, using a temporal convolution neural network that enforces the temporal coherence over the estimated 3D poses. The estimated 3D poses from different viewpoints are then aligned to the same camera reference frame. Finally, we propose to use a temporal convolution network-based classifier for cross-view action recognition. Our results show that we can achieve state of art view-invariant action recognition accuracy even for the challenging viewpoints by only using RGB videos, without pre-training on synthetic or motion capture data. [less ▲]

Detailed reference viewed: 143 (0 UL)
Full Text
Peer Reviewed
See detailThe Impact of Human Mobility on Edge Data Center Deployment in Urban Environments
Vitello, Piergiorgio UL; Capponi, Andrea UL; Fiandrino, Claudio UL et al

in IEEE Global Communications Conference (GLOBECOM), Waikoloa, HI, USA, 2019 (2019, December)

Multi-access Edge Computing (MEC) brings storage and computational capabilities at the edge of the network into so-called Edge Data Centers (EDCs) to better low-latency applications. To this end ... [more ▼]

Multi-access Edge Computing (MEC) brings storage and computational capabilities at the edge of the network into so-called Edge Data Centers (EDCs) to better low-latency applications. To this end, effective placement of EDCs in urban environments is key for proper load balance and to minimize outages. In this paper, we specifically tackle this problem. To fully understand how the computational demand of EDCs varies, it is fundamental to analyze the complex dynamics of cities. Our work takes into account the mobility of citizens and their spatial patterns to estimate the optimal placement of MEC EDCs in urban environments in order to minimize outages. To this end, we propose and compare two heuristics. In particular, we present the mobility-aware deployment algorithm (MDA) that outperforms approaches that do not consider citizens mobility. Simulations are conducted in Luxembourg City by extending the CrowdSenSim simulator and show that efficient EDCs placement significantly reduces outages. [less ▲]

Detailed reference viewed: 66 (8 UL)
See detailTowards a complexity turn in border studies? Rückblicke und Ausblicke
Wille, Christian UL

Presentation (2019, December)

Detailed reference viewed: 113 (0 UL)
Full Text
See detailDas postfaktische Drängen des Buchstaben in Zeiten rechtlich-sprachlicher Entgrenzung
Becker, Katrin UL

in Geisenhanslüke, Achim (Ed.) Buchstäblichkeit Theorie, Geschichte, Übersetzung (2019)

Full Text
Peer Reviewed
See detailFrom tech to bench: Deep Learning pipeline for image segmentation of high-throughput high-content microscopy data
Garcia Santa Cruz, Beatriz UL; Jarazo, Javier UL; Saraiva, Claudia UL et al

Poster (2019, November 29)

Automation of biological image analysis is essential to boost biomedical research. The study of complex diseases such as neurodegenerative diseases calls for big amounts of data to build models towards ... [more ▼]

Automation of biological image analysis is essential to boost biomedical research. The study of complex diseases such as neurodegenerative diseases calls for big amounts of data to build models towards precision medicine. Such data acquisition is feasible in the context of high-throughput screening in which the quality of the results relays on the accuracy of image analysis. Although the state-of-the-art solutions for image segmentation employ deep learning approaches, the high cost of manual data curation is hampering the real use in current biomedical research laboratories. Here, we propose a pipeline that employs deep learning not only to conduct accurate segmentation but also to assist with the creation of high-quality datasets in a less time-consuming solution for the experts. Weakly-labelled datasets are becoming a common alternative as a starting point to develop real-world solutions. Traditional approaches based on classical multimedia signal processing were employed to generate a pipeline specifically optimized for the high-throughput screening images of iPSC fused with rosella biosensor. Such pipeline produced good segmentation results but with several inaccuracies. We employed the weakly-labelled masks produced in this pipeline to train a multiclass semantic segmentation CNN solution based on U-net architecture. Since a strong class imbalance was detected between the classes, we employed a class sensitive cost function: Dice coe!cient. Next, we evaluated the accuracy between the weakly-labelled data and the trained network segmentation using double-blind tests conducted by experts in cell biology with experience in this type of images; as well as traditional metrics to evaluate the quality of the segmentation using manually curated segmentations by cell biology experts. In all the evaluations the prediction of the neural network overcomes the weakly-labelled data quality segmentation. Another big handicap that complicates the use of deep learning solutions in wet lab environments is the lack of user-friendly tools for non-computational experts such as biologists. To complete our solution, we integrated the trained network on a GUI built on MATLAB environment with non-programming requirements for the user. This integration allows conducting semantic segmentation of microscopy images in a few seconds. In addition, thanks to the patch-based approach it can be employed in images with different sizes. Finally, the human-experts can correct the potential inaccuracies of the prediction in a simple interactive way which can be easily stored and employed to re-train the network to improve its accuracy. In conclusion, our solution focuses on two important bottlenecks to translate leading-edge technologies in computer vision to biomedical research: On one hand, the effortless obtention of high-quality datasets with expertise supervision taking advantage of the proven ability of our CNN solution to generalize from weakly-labelled inaccuracies. On the other hand, the ease of use provided by the GUI integration of our solution to both segment images and interact with the predicted output. Overall this approach looks promising for fast adaptability to new scenarios. [less ▲]

Detailed reference viewed: 63 (11 UL)
Full Text
See detailDeep Neural Networks for Personalized Sentiment Analysis with Information Decay
Guo, Siwen UL

Doctoral thesis (2019)

People have different lexical choices when expressing their opinions. Sentiment analysis, as a way to automatically detect and categorize people’s opinions in text, needs to reflect this diversity. In ... [more ▼]

People have different lexical choices when expressing their opinions. Sentiment analysis, as a way to automatically detect and categorize people’s opinions in text, needs to reflect this diversity. In this research, I look beyond the traditional population-level sentiment modeling and leverage socio-psychological theories to incorporate the concept of personalized modeling. In particular, a hierarchical neural network is constructed, which takes related information from a person’s past expressions to provide a better understanding of the sentiment from the expresser’s perspective. Such personalized models can suffer from the data sparsity issue, therefore they are difficult to develop. In this work, this issue is addressed by introducing the user information at the input such that the individuality from each user can be captured without building a model for each user and the network is trained in one process. The evolution of a person’s sentiment over time is another aspect to investigate in personalization. It can be suggested that recent incidents or opinions may have more effect on the person’s current sentiment than the older ones, and the relativeness between the targets of the incidents or opinions plays a role on the effect. Moreover, psychological studies have argued that individual variation exists in how frequently people change their sentiments. In order to study these phenomena in sentiment analysis, an attention mechanism which is reshaped with the Hawkes process is applied on top of a recurrent network for a user-specific design. Furthermore, the modified attention mechanism delivers a functionality in addition to the conventional neural networks, which offers flexibility in modeling information decay for temporal sequences with various time intervals. The developed model targets data from social platforms and Twitter is used as an example. After experimenting with manually and automatically labeled datasets, it can be found that the input formulation for representing the concerned information and the network design are the two major impact factors of the performance. With the proposed model, positive results have been observed which confirm the effectiveness of including user-specific information. The results reciprocally support the psychological theories through the real-world actions observed. The research carried out in this dissertation demonstrates a comprehensive study of the significance of considering individuality in sentiment analysis, which opens up new perspectives for future research in the area and brings opportunities for various applications. [less ▲]

Detailed reference viewed: 128 (4 UL)
Full Text
Peer Reviewed
See detailDreaming of Constructivist Technology Integration Strategies in Future Teacher Students
Reuter, Bob UL; Busana, Gilbert UL

Scientific Conference (2019, November 27)

Based on previous experiences in preparing future teachers for technology integration (Reuter & Busana, 2017), and based on the recommendations from Kolb’s (2017) Triple E framework about effective uses ... [more ▼]

Based on previous experiences in preparing future teachers for technology integration (Reuter & Busana, 2017), and based on the recommendations from Kolb’s (2017) Triple E framework about effective uses of ICT in education, we have adapted the Educational Technology course in our Initial Teacher Training. Over the years, we have indeed observed that, when given the choice of the type of technology integration strategies, many students designed ICT-based learning and teaching scenarios that implemented a rather teacher-centred teaching model (Roblyer & Doering, 2013). These scenarios were often far from innovative nor did they implement the disruptive potential of ICT in education (Christensen, Horn & Johnson, 2008). In the winter semester 2018-2019 we thus decided to ask our students to design and develop constructivist technology integration scenarios. We assessed the success of this adaptation with the help of our own observations, the semester reports produced by our students and their answers to an end-of-semester course evaluation. In general, we saw that students were able to design rather attractive constructivist learning activities. We also observed that our students were quite surprised that such activities do not require complicated and expensive tools, but that they can be implemented with standard productivity tools. [less ▲]

Detailed reference viewed: 19 (4 UL)
Full Text
See detailA professional development course in translanguaging: Challenges and opportunities
Aleksic, Gabrijela UL; Dzoen, Bebic-Crestany

Scientific Conference (2019, November 27)

In Luxembourg, the new law in 2017 has declared multilingual early education mandatory. Not only that teachers need to help children develop their Luxembourgish, but also they need to familiarize them ... [more ▼]

In Luxembourg, the new law in 2017 has declared multilingual early education mandatory. Not only that teachers need to help children develop their Luxembourgish, but also they need to familiarize them with French and value their home languages. In order to support preschool teachers in this endeavour, our project aims to: (1) offer a professional development (PD) course in translanguaging, (2) involve children’s families to reinforce home-school collaboration, and (3) foster children’s cognitive, linguistic, and socio-emotional engagement in the classroom. We use a panoply of measures to reach our aims: focus groups and teacher questionnaires (aim 1), parent questionnaires and interviews (aim 2), a test in early literacy and numeracy in school and home languages, teacher assessment of children’s socio-emotional development and video observations with children (aim 3). Translanguaging, the main topic of our 22 hour PD course (June – December 2019), is the use of a full linguistic repertoire to make meaning (Otheguy, García, & Reid, 2015). In eight sessions, we explore multilingual ecology, parental involvement, and oracy and early literacy. We will present preliminary findings of the focus groups with teachers and tests in early literacy and numeracy in children’s home and school languages. Challenges and opportunities that emerged during the course will be explored as well. [less ▲]

Detailed reference viewed: 50 (1 UL)
Full Text
See detailDe Suchtverband verbënnt - 2. nationales Suchtsymposium, Kommunale Suchtprävention
Böwen, Petra UL

Conference given outside the academic context (2019)

Detailed reference viewed: 8 (0 UL)
Full Text
See detailModeling Parkinson's disease using human midbrain organoids
Monzel, Anna Sophia UL

Doctoral thesis (2019)

With increasing prevalence, neurodegenerative disorders present a major challenge for medical research and public health. Despite years of investigation, significant knowledge gaps exist, which impede the ... [more ▼]

With increasing prevalence, neurodegenerative disorders present a major challenge for medical research and public health. Despite years of investigation, significant knowledge gaps exist, which impede the development of disease-modifying therapies. The development of tools to model both physiological and pathological human brains greatly enhanced our ability to study neurological disorders. Brain organoids, derived from human induced pluripotent stem cells (iPSCs), hold unprecedented promise for biomedical research to unravel novel pathological mechanisms of a multitude of brain disorders. As brain proxies, these models bridge the gap between traditional 2D cell cultures and animal models. Owing to their human origin, hiPSC-derived organoids can recapitulate features that cannot be modeled in animals by virtue of differences in species. Parkinson’s disease (PD) is a human-specific neurodegenerative disorder. The major manifestations are the consequence of degenerating dopaminergic neurons (DANs) in the midbrain. The disease has a multifactorial etiology and a multisystemic pathogenesis and pathophysiology. In this thesis, we used state-of-the-art technologies to develop a human midbrain organoid (hMO) model with a great potential to study PD. hMOs were generated from iPSC-derived neural precursor cells, which were pre-patterned to the midbrain/hindbrain region. hMOs contain multiple midbrain-specific cell types, such as midbrain DANs, as well as astrocytes and oligodendrocytes. We could demonstrate features of neuronal maturation such as myelination, synaptic connections, spontaneous electrophysiological activity and neural network synchronicity. We further developed a neurotoxin-induced PD organoid model and set up a high-content imaging platform coupled with machine learning classification to predict neurotoxicty. Patient-derived hMOs display PD-relevant pathomechanisms, indicative of neurodevelopmental deficits. hMOs as novel in vitro models open up new avenues to unravel PD pathophysiology and are powerful tools in biomedical research. [less ▲]

Detailed reference viewed: 86 (1 UL)
Full Text
See detailLa naturalisation du prince Félix. Un vote très spécial en 1919
Scuto, Denis UL

Article for general public (2019)

Detailed reference viewed: 31 (0 UL)
Full Text
Peer Reviewed
See detailAu-delà du cours magistral: Vers des apprentissages autonomes et collaboratifs basés sur de « grandes questions »
Busana, Gilbert UL; Reuter, Bob UL

Scientific Conference (2019, November 22)

Nous enseignions depuis quelques années un cours de première année sur "l'éducation à l'ère numérique", dans la formation initiale des futur-e-s enseignant-e-s du fondamental, sous forme d’un cours ... [more ▼]

Nous enseignions depuis quelques années un cours de première année sur "l'éducation à l'ère numérique", dans la formation initiale des futur-e-s enseignant-e-s du fondamental, sous forme d’un cours magistral afin de véhiculer certaines connaissances théoriques et pratiques dans ce domaine. Cependant, nous avons observé au cours des dernières années que (1) de nombreux étudiants échouaient à l'examen et que (2) beaucoup d'entre eux, lorsqu'ils étaient en 4e année, avaient oublié la plupart des fondements. Ces considérations nous ont amenés à repenser et à remanier nos méthodes d'enseignement et notre façon d'évaluer les résultats d'apprentissage des étudiant-e-s. Après une première implémentation de notre nouvelle approche pédagogique combinant diverses méthodes actives d'apprentissage au semestre d’hiver 2018 (décrite et réfléchie dans Reuter & Busana, 2018), nous avons effectué quelques modifications basées (a) sur nos propres observations et réflexions ainsi que (b) sur les commentaires des étudiant-e-s tirées de l’évaluation du cours (organisée par notre université). Dès à présent, plutôt que de les faire travailler sur de « grandes questions » (c.-à-d. des questions pour lesquelles il n'y a pas de réponses simples et faciles à consulter) au cours des séances en présentiel, nous demandons maintenant à nos étudiant-e-s (1) de construire individuellement des connaissances en amont du séminaire sur base de ressources fournies ou tirés de recherches bibliographiques autonomes ; (2) de les présenter, discuter et développer de manière collaborative en petits groupes pendant les séminaires ; puis (3) de les partager et discuter avec nous et leurs pairs et finalement (4) de les conserver dans un portfolio numérique personnel. L’évaluation des apprentissages des étudiant-e-s est effectuée à l’aide de ce portfolio numérique en fin de semestre. [less ▲]

Detailed reference viewed: 75 (4 UL)
Peer Reviewed
See detailDie Entstehung einer Hauptstadtregion als deutsch-belgische Beziehungsgeschichte
Brüll, Christoph UL

Scientific Conference (2019, November 22)

Vom 14. Mai - 23. Juli und vom 19. September - 31.Oktober 1972 fand zunächst im Kölner Museum Schnütgen und dann in den Brüsseler Musées Royaux d’Art et d’Histoire die bis heute größte Schau rheinisch ... [more ▼]

Vom 14. Mai - 23. Juli und vom 19. September - 31.Oktober 1972 fand zunächst im Kölner Museum Schnütgen und dann in den Brüsseler Musées Royaux d’Art et d’Histoire die bis heute größte Schau rheinisch-maasländischer Kunst statt. Unter dem Titel „Rhein und Maas. Kunst und Kultur 800-1400“ wurde sie zu einem großen Publikumserfolg, der auch in der Scientific Community der Kunsthistoriker einige Aufmerksamkeit erregte. Die Entstehung der Ausstellung sagt einiges über die Geschichte der Wissenschaftsbeziehungen zwischen Deutschland und Belgien seit 1945 aus, aber auch, so die Ausgangshypothese meines Beitrags, über die Veränderungen der in Belgien vorherrschenden Mental Maps von Deutschland seit der Verlagerung des politischen Schwerpunktes an den Rhein mit der Gründung der Bundesrepublik. Der belgische Deutschlanddiskurs knüpfte dabei durchaus an bereits bestehende Deutschlandbilder an, die gerade das Maasland als Zwischenraum inszenierten , doch kam durch das Zusammenrücken der westeuropäischen Länder im Zuge von Kaltem Krieg und beginnender europäischer Integration, gerade auch unter dem Signum des Abendlandes, eine neue Dimension ins Spiel. Mentalitätsgeschichtlich hob die rheinische Bonner Republik die durch den Ersten Weltkrieg hervorgerufene Distanz zwischen dem belgischen Staat und (West-)Deutschland auf, politikgeschichtlich fanden deutsche und belgische „Suche nach Sicherheit“ (E. Conze) ein enges Kooperationsfeld. Zum Zentrum der belgischen Aktivität im Rheinland wurde Köln, das nach dem durch die Wahl als Regierungssitz notwendig gewordenen Verlassen Bonns durch die belgischen Truppen in Deutschland 1950 zu deren neuem Hauptquartier wurde. Dort etablierte sich zudem rasch das Belgische Haus als Drehscheibe für politischen, konsularischen, aber auch kulturellen Austausch. Die Ausstellung „Rhein und Maas“ dient als Aufhänger für den Beitrag, der sowohl das Ende einer langen Nachkriegszeit als auch das Ankommen Belgiens in der Bonner Republik herausstellen kann. Auch die Anfang der 1970er Jahre erfolgten ersten gegenseitigen Staatsbesuche geben Aufschluss über Inszenierung und Wahrnehmung der Hauptstadtregion. Der Beitrag wirft somit eine beziehungsgeschichtliche Perspektive auf Bonn-Köln-Düsseldorf zwischen 1949 und den 1970er Jahren und fragt mit Bezug auf die Hauptstadtregion auf die belgische(n) Wahrnehmung(en) des Rheinischen, aber auch nach Rückwirkungen auf Brüssel als belgischer und europäischer Hauptstadt. [less ▲]

Detailed reference viewed: 69 (0 UL)
See detailEinschnitte mit Langzeitwirkung: Kriegsende und Nachkriegszeit in Ostbelgien (1944-1950)
Brüll, Christoph UL

Conference given outside the academic context (2019)

Detailed reference viewed: 74 (0 UL)
See detailL'administration des postes et des télégraphes : une institution entre responsabilité internationale et intérêts nationaux
Lafontaine, Aurélia UL

Presentation (2019, November 20)

Dans la deuxième moitié du XIXe siècle, le Grand-Duché de Luxembourg devient membre de deux organismes multilatéraux des (télé)communications, à savoir l’Union télégraphique internationale (UIT) et ... [more ▼]

Dans la deuxième moitié du XIXe siècle, le Grand-Duché de Luxembourg devient membre de deux organismes multilatéraux des (télé)communications, à savoir l’Union télégraphique internationale (UIT) et l’Union postale générale (UPU). Y siègent, non pas des diplomates, peu actifs dans un pays soumis, depuis le traité de Londres de 1867, à un statut de neutralité perpétuelle et désarmée, mais des représentants de l’administration des postes et des télégraphes. S’inscrivant dans une tendance historiographique récente et axée sur les concepts de la techno- diplomatie et de l’expert, l’exposé propose un cadre théorique combinant des notions de la théorie des petites puissances, issue des sciences politiques, et des concepts de sociologie. Sur cette base et pour la période comprise entre la naissance de l’UIT en 1865 et le début de la Première Guerre mondiale en 1914, la présentation vise à fournir des éléments de réponse à une question de recherche double. D’une part, il s’agit d’étudier le rôle et la stratégie du Grand-Duché et notamment des experts de l’administration des postes et des télégraphes au sein des organismes multilatéraux. Dans ce cadre, il convient également d’analyser si la définition de l’expert, telle que proposée dans la littérature scientifique, peut être appliquée aux représentants luxembourgeois ou si le Grand-Duché représente un cas particulier. D’autre part, l’exposé entend mettre en évidence les dynamiques de coopération, mais aussi les tensions qui surgissent dans ce contexte entre les trois institutions grand-ducales impliquées dans la représentation du pays, à savoir l’administration des postes, son ministère de tutelle qui est la Direction générale des Finances, ainsi que le Ministère des Affaires étrangères. Enfin, les corollaires de la représentation du Grand-Duché à l’UIT et à l’UPU seront étudiés. Il s’agira en particulier de déterminer si et comment le Luxembourg adapte son attitude dans d’autres formes de relations internationales en rapport avec les (télé)communications suite à l’émergence des unions et des standards qu’elles imposent. [less ▲]

Detailed reference viewed: 24 (0 UL)
Full Text
Peer Reviewed
See detailOn degree-d zero-sum sets of full rank
Beierle, Christof UL; Biryukov, Alex UL; Udovenko, Aleksei UL

in Cryptography and Communications (2019)

A set 𝑆⊆𝔽𝑛2 is called degree-d zero-sum if the sum ∑𝑠∈𝑆𝑓(𝑠) vanishes for all n-bit Boolean functions of algebraic degree at most d. Those sets correspond to the supports of the n-bit Boolean ... [more ▼]

A set 𝑆⊆𝔽𝑛2 is called degree-d zero-sum if the sum ∑𝑠∈𝑆𝑓(𝑠) vanishes for all n-bit Boolean functions of algebraic degree at most d. Those sets correspond to the supports of the n-bit Boolean functions of degree at most n − d − 1. We prove some results on the existence of degree-d zero-sum sets of full rank, i.e., those that contain n linearly independent elements, and show relations to degree-1 annihilator spaces of Boolean functions and semi-orthogonal matrices. We are particularly interested in the smallest of such sets and prove bounds on the minimum number of elements in a degree-d zero-sum set of rank n. The motivation for studying those objects comes from the fact that degree-d zero-sum sets of full rank can be used to build linear mappings that preserve special kinds of nonlinear invariants, similar to those obtained from orthogonal matrices and exploited by Todo, Leander and Sasaki for breaking the block ciphers Midori, Scream and iScream. [less ▲]

Detailed reference viewed: 36 (0 UL)
Full Text
Peer Reviewed
See detailParliamentary Administrations in EU Politics: The Sources and Impact of Diversity
Högenauer, Anna-Lena UL

E-print/Working paper (2019)

National parliaments tend to delegate a range of tasks related to EU affairs scrutiny to their administrations. However, there is considerable variation in the extent to which national parliaments invest ... [more ▼]

National parliaments tend to delegate a range of tasks related to EU affairs scrutiny to their administrations. However, there is considerable variation in the extent to which national parliaments invest in EU affairs staff. This raises two questions: what is the source of this diversity, and what is its impact on the EU affairs scrutiny of national parliaments? This paper argues that greater investments in EU affairs staff are associated with public Euroscepticism, large parliaments and parliaments with strong scrutiny powers. A large EU staff is, in turn, associated with parliaments that are relatively active in EU affairs scrutiny. This is particularly true for the 15 EU member states that joined before 2004. [less ▲]

Detailed reference viewed: 14 (0 UL)