Search

The security of vehicle communication buses and electronic control units has received much attention in the recent years. However, while essential for practical deployments, the problem of securely exchanging cryptographic keys between electronic control units on the CAN bus received little attention so far. In this work, we evaluate group extensions of a regular key exchange protocol, i.e., the elliptic curve version of the Diffie-Hellman protocol, by using both a standardized NIST elliptic curve as well as the faster, more recently proposed Four $\mathbb {Q}$ curve. We deploy protocol implementations and determine crisp performance bounds on real-world automotive-grade platforms with Infineon and ARM cores. For an up-to-date analysis, we use both CAN and its more recent extension CAN-FD as communication layers. Roughly, the computational runtime of the key exchange protocol scales logarithmically or linearly with the number of nodes, depending on the protocol version. The computational time proves to be more critical than bandwidth due to the more demanding elliptic curve operations.

Background Eating disorders (ED) involve both the nervous system and the gastrointestinal tract. A similar double involvement is also found in disorders of the brain-gut interaction (DGBI) and symptoms are sometimes similar. Aim To find out where there is an association and a cause-effect relationship, we looked for the comorbidity of DGBI and ED. Methods A systematic review was undertaken. A literature search was performed. Inclusion criteria for the articles retained for analysis were: Observational cohort population-based or hospital-based and case-control studies, examining the relationship between DGBI and ED. Exclusion criteria were: Studies written in other languages than English, abstracts, conference presentations, letters to the Editor and editorials. Selected papers by two independent investigators were critically evaluated and included in this review. Results We found 29 articles analyzing the relation between DGBI and ED comprising 13 articles on gastroparesis, 5 articles on functional dyspepsia, 7 articles about functional constipation and 4 articles on irritable bowel syndrome. Conclusion There is no evidence for a cause-effect relationship between DGBI and ED. Their common symptomatology requires correct identification and a tailored therapy of each disorder.

The Controller Area Network (CAN) is the most widely employed communication protocol for in-vehicle applications. While many of its features qualify it as a suitable candidate for future use in automotive networking, the lack of security mechanisms makes it problematic for safety-critical applications. Recently, both the research community and the industry have proposed a large number of solutions for securing CAN, but most of these solutions put additional strain on the already limited 8 byte CAN payload or require more expensive hardware. In this work, we propose the use of a covert voltage channel that can be used for the transmission of additional data required by specific security mechanisms. We achieve this with the help of additional transceivers by encoding additional bits as different voltage levels in the existing CAN dominant bits without affecting regular CAN traffic decoding. We demonstrate the application of our approach on both low-end and high-end automotive embedded platforms and prove its suitability for implementing authentication mechanisms and key exchange protocols over CAN while maintaining backward compatibility.

Background and Aims. Both nonalcoholic fatty liver disease (NAFLD) and ischemic heart disease have common pathogenic links. Evidence for the association of NAFLD with acute coronary syndromes (ACS), complex multivessel coronary artery disease (CAD), and increased mortality risk in ACS patients is still under investigation. Therefore, we conducted a systematic review aiming to clarify these gaps in evidence. Methods. We conducted a systematic search on PubMed and EMBASE with predefined keywords searching for observational studies published till August 2020. NAFLD diagnosis was accepted if confirmed through biopsy, imaging techniques, surrogate markers, or codes. Full articles that satisfied our inclusion and exclusion criteria were included in the systematic review. We used the NHLBI quality assessment tool to evaluate included studies. Results. Seventeen observational studies with a total study population of approximately 21 million subjects were included. Eleven studies evaluated whether NAFLD is an independent risk factor for developing ACS with conflicting results, of which eight studies demonstrated a significant association between NAFLD and ACS, mainly in Asian populations, while three reported a lack of an independent association. Conflicting results were reported in studies conducted in Europe and North America. Moreover, a total of five studies evaluated whether NAFLD and fatty liver severity in ACS patients are associated with a complex multivessel CAD disease, where all studies confirmed a significant association. Furthermore, seven out of eight studies evaluating NAFLD and hepatic steatosis severity as a predictor of all-cause and cardiovascular mortality and in-hospital major adverse cardiovascular events (MACE) in ACS patients demonstrated a significant independent association. Conclusions. NAFLD patients are associated with an independently increased risk of developing ACS, mainly in Asian populations, with inconsistent results in North American and European individuals. Moreover, NAFLD and hepatic steatosis severity were both independently correlated with complex multivessel CAD, mortality, and in-hospital MACE in ACS patients.

Objectives. Continuous delivery of levodopa-carbidopa intestinal gel (LCIG) represents a gold standard option for the treatment of advanced PD and the benefits of intestinal gel delivery are significative superior from those obtained with available medical therapies. We aimed to investigate the percutaneous endoscopic transgastric jejunostomy (PEG-J) use and the occurrence of procedural complications in a group of patients diagnosed with PD Hoehn and Jahr 3 and 4 stages, receiving LCIG therapy. Materials and methods. We performed an observational retrospective cohort study on 24 PD patients recruited from a single regional tertiary center in Cluj-Napoca, Romania, between 2009 and 2019. We included PD patients with Hoehn and Jahr 3 and 4 stages under LCIG therapy.results. PEG-J replacement was found in 54.2% of the patients. Erythema and inflammation at the level of stoma (37.5%) and allergic reactions (20.8%) were the most common complications. The most frequent reason for PEG-J replacement was tube dislocation (20.8%), and rarely the plied tube, the excess of adipose tissue, or bezoar (4.3% each). A third of the patients received PEG-J therapy during the night, after a median time of three years, after PEG-J initiation. A quarter of all the subjects had their Levodopa-Carbidopa therapy interrupted. The median time untill interruption was one year. The reasons for the interruption were patient, or patient’s relative refuses, for a sixth of the cases, or gastric polyps, in very few cases.conclusions. The most frequent complication of LCIG delivered by PEG-J was erythema and inflammation at the level of stoma. However LCIG therapy discontinuation was in particular due to the decision of the patient or relatives.

Most of the existing works on securing the CAN bus are using the limited data-field of CAN frames to embed a cryptographic payload. Only very few works have suggested the use of the identifier field since identifiers are critical for the arbitration procedure and changing them at random would interfere with message priorities. To preserve priority on the bus, in this work we use an ordered CMAC buffer. In this way, we can authenticate the identifiers of CAN frames and check that the sender is a legitimate node while arbitration on the bus remains unaltered. Moreover, we determine that for real-world scenarios the achieved security level is very close to the length of the ID field despite the constraints from ordering. This procedure easily circumvents replay attacks and fuzz testing on the bus, which were exploited by many recent works. We prove the feasibility of our approach by testing practical implementations on automotive-grade microcontrollers and CAN-bus traffic allocations from a high-end vehicle. The computational requirements are some of the lowest achievable for securing CAN, with a dozen CMAC-AES computations being sufficient for extracting a table of one hundred identifiers.

Background Non-alcoholic steatohepatitis (NASH) is characterized by hepatic steatosis with inflammation, ballooned hepatocytes, and possible fibrosis, which may progress to liver cirrhosis. Although liver biopsy, remains the diagnostic gold standard of NASH, several noninvasive biomarkers have been studied, to avoid the need for this invasive procedure. We performed a systematic review with meta-analysis to evaluate the accuracy of several noninvasive biomarkers in predicting NASH and assessing liver fibrosis in NASH patients. Evidence acquisition An electronic search on PubMed and EMBASE was systematically performed. The principal summary outcome was the area under the curve (AUC), assessing the accuracy of NashTest, BARD (body mass index, AST/ALT ratio, diabetes) score, NAFLD fibrosis score (NFS), APRI (aspartate aminotransferase-to-platelet ratio index), and fibrosis-4 (FIB-4) index in predicting NASH and assessing liver fibrosis. Evidence synthesis Thirteen studies involving 6,557 adult patients were included in the qualitative assessment of this review, out of which, six studies were included in the quantitative assessment. Prediction of NASH was evaluated better using NFS (AUC of 0.687) and FIB-4 (AUC of 0.729). Fibrosis stages 0 vs. 1-4 was diagnosed better using NFS (AUC of 0.718) and FIB-4 (AUC of 0.723). Advanced fibrosis was assessed better by BARD (AUC of 0.673), APRI (AUC of 0.762), NFS (AUC of 0.787) and FIB-4 (AUC of 0.821). Conclusions FIB-4 predicted NASH and quantified liver fibrosis, stages 0 vs. 1-4 more precisely compared to NFS, APRI, and BARD. However, considering that methodological quality of the assessed studies is limited, the results should be considered with caution.

(1) Background: The role of adipokines such as adiponectin and visfatin in metabolic-dysfunction-associated fatty liver disease (MAFLD) and cardiovascular disease remains unclear. Therefore, we aim to assess serum adiponectin and visfatin levels in MAFLD patients and associated cardiovascular parameters. (2) Methods: A cross-sectional study involving 80 participants (40 MAFLD patients, 40 controls), recruited between January and September 2020, was conducted, using both hepatic ultrasonography and SteatoTestTM to evaluate hepatic steatosis. Echocardiographic and Doppler parameters were assessed. Serum adipokines were measured using ELISA kits. (3) Results: Adiponectin and visfatin levels were not significantly different in MAFLD vs. controls. Visfatin was associated with mean carotid intima-media thickness (p-value = 0.047), while adiponectin was associated with left ventricular ejection fraction (LVEF) (p-value = 0.039) and E/A ratio (p-value = 0.002) in controls. The association between adiponectin and E/A ratio was significant in the univariate analysis at 95% CI (0.0049–0.1331, p-value = 0.035), but lost significance after the multivariate analysis. Although LVEF was not associated with adiponectin in the univariate analysis, significant values were observed after the multivariate analysis (95% CI (−1.83–−0.22, p-value = 0.015)). (4) Conclusions: No significant difference in serum adiponectin and visfatin levels in MAFLD patients vs. controls was found. Interestingly, although adiponectin levels were not associated with LVEF in the univariate analysis, a significant inversely proportional association was observed after the multivariate analysis.

Background Functional dyspepsia (FD) is one of the most frequent functional gastrointestinal disorders and is defined using the Rome IV criteria as any combination of the following symptoms: postprandial fullness, early satiety, epigastric pain, and epigastric burning that are severe enough to interfere with the usual activities and occur at least 3 days per week over the past 3 months with an onset of at least 6 months before the presentation. The purpose of this systematic review is to analyze all the relevant studies in the literature that investigate the efficiency of hypnotherapy in FD. Areas of uncertainty FD refractory to conservative treatment is a therapeutic challenge, and alternative treatment options are needed. Gut-oriented hypnotherapy has been reported an effective treatment for irritable bowel syndrome, but poorly tested in FD. Data sources We performed a search in 6 bibliographic databases (PubMed, Embase, Cochrane Library, Web of Science, Scopus, and LILACS) using customized search strategies for each engine. The search strategy included the following terms: (hypnosis, hypnotherapy, hypnotherapies, hypnogenesis, hypnotism, hypnotist, hypnotical suggestion, suggestion, and mesmerism) and {[functional and (dyspepsia or dyspeptic)] or FD}. Results Taking the aforementioned criteria into account, the result was a review of 4 articles analyzing the efficacy of hypnotherapy in the treatment of FD, published in the past 20 years. The initial search identified 398 articles, of which 37 potentially appropriate articles were reviewed. Of these 37 articles, 4 articles were included in the review. The benefits observed by numerous studies go beyond the field of digestive pathology, patients describing a general improvement in physical and mental health. Conclusions Current studies analyzing the efficacy of hypnotherapy in FD provide encouraging data, but additional randomized controlled trials are needed before a firm position on the effectiveness of hypnosis in FD.

This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable. Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability in NLP. Then, we will discuss explainability for NLP tasks and report on a systematic literature review of the state-of-the-art literature in AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data.

The interpretability or explainability of AI systems (XAI) has been a topic gaining renewed attention in recent years across AI and HCI communities. Recent work has drawn attention to the emergent explainability requirements of in situ, applied projects, yet further exploratory work is needed to more fully understand this space. This paper investigates applied AI projects and reports on a qualitative interview study of individuals working on AI projects at a large technology and consulting company. Presenting an empirical understanding of the range of stakeholders in industrial AI projects, this paper also draws out the emergent explainability practices that arise as these projects unfold, highlighting the range of explanation audiences (who), as well as how their explainability needs evolve across the AI project lifecycle (when). We discuss the importance of adopting a sociotechnical lens in designing AI systems, noting how the “AI lifecycle” can serve as a design metaphor to further the XAI design field.

Entity linking (EL), the task of disambiguating mentions in text by linking them to entities in a knowledge graph, is crucial for text understanding, question answering or conversational systems. Entity linking on short text (e.g., single sentence or question) poses particular challenges due to limited context. While prior approaches use either heuristics or black-box neural methods, here we propose LNN-EL, a neuro-symbolic approach that combines the advantages of using interpretable rules based on first-order logic with the performance of neural learning. Even though constrained to using rules, LNN-EL performs competitively against SotA black-box neural approaches, with the added benefits of extensibility and transferability. In particular, we show that we can easily blend existing rule templates given by a human expert, with multiple types of features (priors, BERT encodings, box embeddings, etc), and even scores resulting from previous EL methods, thus improving on such methods. For instance, on the LC-QuAD-1.0 dataset, we show more than $4$\% increase in F1 score over previous SotA. Finally, we show that the inductive bias offered by using logic results in learned rules that transfer well across datasets, even without fine tuning, while maintaining high accuracy., Accepted to ACL 2021

Background and Aims: The nonpharmacological therapy in irritable bowel syndrome (IBS) is expanding rapidly. Practitioners and medical educators need to be aware of progress and changes in knowledge of this topic. The Romanian Society of Neurogastroenterology aimed to create guidelines based on best evidence on the use of nonpharmacological therapy in IBS. Methods: A group of experts was constituted. This was divided in eleven subgroups dedicated to eleven categories of nonpharmacological therapy. The subgroups searched the literature and formulated statements and recommendations. These were submitted to vote in order to obtain consensus. Results: The outcome of this activity is represented by the guidelines of the Romanian Society of Neurogastroenterology, presented in this paper. The recommendations are seen as complementary to the pharmacological therapy and are not intended to recommend avoiding pharmacological drugs. Conclusions: These guidelines were elaborated by a Delphi process and represent a useful tool for physicians managing patients with IBS.

We present XNLP: an interactive browser-based system embodying a living survey of recent state-of-the-art research in the field of Explainable AI (XAI) within the domain of Natural Language Processing (NLP). The system visually organizes and illustrates XAI-NLP publications and distills their content to allow users to gain insights, generate ideas, and explore the field. We hope that XNLP can become a leading demonstrative example of a living survey, balancing the depth and quality of a traditional well-constructed survey paper with the collaborative dynamism of a widely available interactive tool. XNLP can be accessed at: https://xainlp2020.github.io/xainlp.

Knowledge base question answering (KBQA)is an important task in Natural Language Processing. Existing approaches face significant challenges including complex question understanding, necessity for reasoning, and lack of large end-to-end training datasets. In this work, we propose Neuro-Symbolic Question Answering (NSQA), a modular KBQA system, that leverages (1) Abstract Meaning Representation (AMR) parses for task-independent question understanding; (2) a simple yet effective graph transformation approach to convert AMR parses into candidate logical queries that are aligned to the KB; (3) a pipeline-based approach which integrates multiple, reusable modules that are trained specifically for their individual tasks (semantic parser, entity andrelationship linkers, and neuro-symbolic reasoner) and do not require end-to-end training data. NSQA achieves state-of-the-art performance on two prominent KBQA datasets based on DBpedia (QALD-9 and LC-QuAD1.0). Furthermore, our analysis emphasizes that AMR is a powerful tool for KBQA systems., Accepted to Findings of ACL

Functional anorectal pain syndromes are a neglected yet often disabling clinical entity resulting in significant economic and psychological burden to the patient. The aim of this review is to update the practicing gastroenterologist/coloproctologist on the diagnosis and management of these complicated disorders. The updated Rome foundation diagnostic criteria (Rome IV) for functional anorectal pain subgroups chronic proctalgia (levator ani syndrome and unspecified functional anorectal pain) and acute proctalgia (proctalgia fugax) on the basis of symptom duration and digital rectal examination findings. Chronic proctalgia is thought to be secondary to paradoxical pelvic floor contraction in many patients and biofeedback to improve the defecation effort has proven effective for over 90% in the short term. Unfortunately, management of proctalgia fugax remains challenging and treatment outcomes modest at best. A number of therapies to relax the pelvic floor may be employed to improve symptoms in functional anorectal pain syndromes; however, only biofeedback to improve defaecatory dynamics in patients with levator ani syndrome has proven effectiveness in a randomized setting. Further investigation of treatment approaches in proctalgia fugax is required.

There are dozens of proposals for securing the controller area network (CAN); however, only a few of them are concerned on how to share secret keys between CAN nodes. Recently, some works have used the non-destructive property of CAN arbitration in order to exchange a secret key and achieve information theoretic security for the key exchange. In our proposals, we exploit both delays and the non-destructive arbitration of CAN to achieve a secure key exchange. While our approach is less efficient when it comes to bandwidth, we do not require any kind of additional hardware and we build our implementation on the software layer which is accessible for any CAN-based application. To boost efficiency, we finally bootstrap secret keys by means of the guessing-resilient protocols, such as encrypted-key-exchange (EKE) and simple password exponential key exchange (SPEKE). In principle, a few dozen frames suffice for a secure key-exchange between two CAN nodes. We discuss several protocol versions and extensions for the case of more than two parties. We also present the experimental results on modern automotive-grade controllers to prove the performance of our solution.

Entity name disambiguation is an important task for many text-based AI tasks. Entity names usually have internal semantic structures that are useful for resolving different variations of the same entity. We present, PARTNER, a deep learning-based interactive system for entity name understanding. Powered by effective active learning and weak supervision, PARTNER can learn deep learning-based models for identifying entity name structure with low human effort. PARTNER also allows the user to design complex normalization and variant generation functions without coding skills.

Entity Matching (EM) is a core data cleaning task, aiming to identify different mentions of the same real-world entity. Active learning is one way to address the challenge of scarce labeled data in practice, by dynamically collecting the necessary examples to be labeled by an Oracle and refining the learned model (classifier) upon them. In this paper, we build a unified active learning benchmark framework for EM that allows users to easily combine different learning algorithms with applicable example selection algorithms. The goal of the framework is to enable concrete guidelines for practitioners as to what active learning combinations will work well for EM. Towards this, we perform comprehensive experiments on publicly available EM datasets from product and publication domains to evaluate active learning methods, using a variety of metrics including EM quality, #labels and example selection latencies. Our most surprising result finds that active learning with fewer labels can learn a classifier of comparable quality as supervised learning. In fact, for several of the datasets, we show that there is an active learning combination that beats the state-of-the-art supervised learning result. Our framework also includes novel optimizations that improve the quality of the learned model by roughly 9% in terms of F1-score and reduce example selection latencies by up to 10x without affecting the quality of the model., accepted for publication in ACM-SIGMOD 2020, 15 pages

Functional gastrointestinal disorders (FGIDs) are a highly prevalent group of heterogeneous disorders, and their diagnostic criteria are symptom-based, with the absence of anatomical and biochemical abnormalities of the gastrointestinal tract. Chronic visceral symptoms are common both in patients with an identifiable organic disease but also in FGID patients. Patients suffering from upper gastrointestinal functional disorders typically present with various symptoms such as early satiety, postprandial fullness, bloating, nausea, vomiting, and epigastric pain. Considering their increasing prevalence, difficulties in diagnosis, and low quality of life, FGIDs have become an emerging problem in gastroenterology. We aimed to provide an updated summary of pathways involved in visceral sensitization. We examined the recent literature searching for evidence of the most important studies about the mechanisms underlying gastrointestinal symptom generation and sensitization.

Documents are often used for knowledge sharing and preservation in business and science, within which are tables that capture most of the critical data. Unfortunately, most documents are stored and distributed as PDF or scanned images, which fail to preserve logical table structure. Recent vision-based deep learning approaches have been proposed to address this gap, but most still cannot achieve state-of-the-art results. We present Global Table Extractor (GTE), a vision-guided systematic framework for joint table detection and cell structured recognition, which could be built on top of any object detection model. With GTE-Table, we invent a new penalty based on the natural cell containment constraint of tables to train our table network aided by cell location predictions. GTE-Cell is a new hierarchical cell detection network that leverages table styles. Further, we design a method to automatically label table and cell structure in existing documents to cheaply create a large corpus of training and test data. We use this to enhance PubTabNet with cell labels and create FinTabNet, real-world and complex scientific and financial datasets with detailed table structure annotations to help train and test structure recognition. Our framework surpasses previous state-of-the-art results on the ICDAR 2013 and ICDAR 2019 table competition in both table detection and cell structure recognition with a significant 5.8% improvement in the full table extraction system. Further experiments demonstrate a greater than 45% improvement in cell structure recognition when compared to a vanilla RetinaNet object detection model in our new out-of-domain FinTabNet.

Information extraction (IE) refers to the task of turning text documents into a structured form, in order to make the information contained therein automatically processable. Ontology Mediated Information Extraction (OMIE) is a new paradigm for IE that seeks to exploit the semantic knowledge expressed in ontologies to improve query answering over unstructured data (properly raw text). In this paper we present Mastro System-T, an OMIE tool born from a joint collaboration between the University of Rome "La Sapienza" and IBM Research Almaden and its first application in a financial domain, namely to facilitate the access to and the sharing of data extracted from the EDGAR system.

Structured representations of entity names are useful for many entity-related tasks such as entity normalization and variant generation. Learning the implicit structured representations of entity names without context and external knowledge is particularly challenging. In this paper, we present a novel learning framework that combines active learning and weak supervision to solve this problem. Our experimental evaluation show that this framework enables the learning of high-quality models from merely a dozen or so labeled examples.

Research efforts related to in-vehicle communication security were largely focused on the Controller Area Network (CAN) protocol. While CAN is still the most widely used protocol for building in-vehicle networks, many safety critical functionalities are based on other communication protocols such as FlexRay or Ethernet which constantly expand their use inside vehicles. In this paper we address the problem of authenticating transmissions in FlexRay networks. We approach this task by adapting an authentication protocol to the time-triggered nature of FlexRay communication while also accounting for non-deterministic transmissions that may occur in the FlexRay dynamic segment. We illustrate the effects of introducing authentication on keeping strict message deadlines by evaluating our proposal based on a real-life scenario from a major vehicle manufacturer.

This tutorial is intended for researchers and practitioners working in the data integration area and, in particular, entity resolution (ER), which is a sub-area focused on linking entities across heterogeneous datasets. We outline the ideal requirements of modern ER systems: (1) capture domain knowledge via (minimal) human interaction, (2) provide as much automation as possible via machine learning techniques, and (3) achieve high explainability. We describe recent research trends towards bringing such ideal ER systems closer to reality. We begin with an overview of human-in-the-loop methods that are based on techniques such as crowdsourcing and active learning. We then dive into recent trends that involve deep learning techniques such as representation learning to automate feature engineering, and combinations of transfer and active learning to reduce the amount of user labels required. We also discuss how explainable AI relates to ER, and outline some of the recent advances towards explainable ER.

As cryptography is quickly entering the automotive domain, public-key cryptographic functions are a vital building block and are part of recent industry-proposed standards. Elliptic curves provide a more compact representation for public/private keys making them more suitable for embedded devices with limited amounts of memory. Nonetheless, they provide more compact signatures and open road for identity-based cryptographic primitives by exploiting the flexibility of bilinear pairings. In this work we carry a performance evaluation on some modern libraries, e.g., MIRACL, RELIC, and compare them to the more classical WolfSSL. The evaluation is carried on a state-of-the-art representative controller from the automotive industry, i.e., a 32 bit Infineon TC297. Having a crisper image on computational requirements is relevant for future automotive and industrial applications.

Entity resolution (ER) is the task of identifying different representations of the same real-world entities across databases. It is a key step for knowledge base creation and text mining. Recent adaptation of deep learning methods for ER mitigates the need for dataset-specific feature engineering by constructing distributed representations of entity records. While these methods achieve state-of-the-art performance over benchmark data, they require large amounts of labeled data, which are typically unavailable in realistic ER applications. In this paper, we develop a deep learning-based method that targets low-resource settings for ER through a novel combination of transfer learning and active learning. We design an architecture that allows us to learn a transferable model from a high-resource setting to a low-resource one. To further adapt to the target dataset, we incorporate active learning that carefully selects a few informative examples to fine-tune the transferred model. Empirical evaluation demonstrates that our method achieves comparable, if not better, performance compared to state-of-the-art learning-based methods while using an order of magnitude fewer labels., This paper is accepted by ACL 2019

(1) Background: Recently, adipokines, including visfatin, have been studied in nonalcoholic fatty liver disease (NAFLD). Several studies evaluated visfatin levels in NAFLD, the presence and severity of hepatic steatosis, liver fibrosis, lobar inflammation, nonalcoholic steatohepatitis (NASH), and gender differences. However, inconclusive results have been reported. Accordingly, we performed a systematic review and meta-analysis, aiming to address these gaps in evidence. (2) Methods: We performed a systematic electronic search on PubMed, EMBASE, and Cochrane Library using predefined keywords. Diagnosis of NAFLD by liver biopsy or imagistic investigations was accepted. Full articles satisfying our inclusion and exclusion criteria were included. NHLBI quality assessment tools were used to evaluate included studies. The principal summary outcome was the mean difference in visfatin levels. (3) Results: There were 21 studies involving 1923 individuals included in our qualitative assessment, while 14 studies were included in the quantitative assessment. No statistical significance was found assessing visfatin levels in NAFLD [3.361 (95% CI −0.175–6.897)], simple steatosis [7.523 (95% CI −16.221–31.267)], hepatic steatosis severity [−0.279 (95% CI −1.843–1.285)], liver fibrosis [4.133 (95% CI −3.176–11.443)], lobar inflammation [0.358 (95% CI −1.470–2.185)], NASH [−2.038 (95% CI −6.839–2.763)], and gender [(95% CI −0.554–0.556)]. (4) Conclusions: In conclusion, visfatin levels are not associated with NAFLD, presence or severity of hepatic steatosis, liver fibrosis, lobar inflammation, NASH, and gender. However, due to the limited methodological quality of the included studies, results should be interpreted with caution.

Levodopa-carbidopa intestinal gel (LCIG) is a method of continuous administration of levodopa - the standard treatment in Parkinson disease (PD, a neurodegenerative disorder characterized by resting tremor, rigidity, gait impairment, and bradykinesia), thought to reduce the short-life and pulsatile problems of oral administration. We aimed to study the effects of Levodopa-Carbidopa therapy in 2 separate groups: one with intrajejunal administration of Levodopa-Carbidopa gel and the second with oral therapy.We performed an observational retrospective Romanian cohort study on 61 patients diagnosed with PD patients, with Hoehn and Jahr 3 and 4 stages, recruited from a single regional tertiary center in Cluj-Napoca, Romania, between 2009 and 2019.The mean adjusted UPDRS III (and similarly for UPDRS II) improved in the LCIG compared to the oral therapy group with 15.6 (95% CI 12.0-19.2, P < .001), and with 18.4 (95% CI 13.8-22.9, P < .001), stratified for the Hoehn and Jahr stages 3 and 4. There was a 41.7% (10) reduction in dyskinesia, and 29.2% reduction in wearing off/on-off at 1 year in the LCIG group compared to 0% (0) dyskinesia reduction, and 2.7% reduction in wearing off/on-off in the oral therapy group.Continuous intrajejunal infusion of LCIG ensures a significant and clinical reduction in motor fluctuations compared to oral therapy in advanced PD, even after adjustment for important confounders.

Car sharing and car access control from mobile devices is an increasingly relevant topic. While numerous proposals started to appear, practical deployments ask for simple solutions, that are easy to implement and yet secure. In this work we explore the use of TPM 2.0 functionalities along with identity-based signatures in order to derive a flexible solution for gaining access to a vehicle. While TPM 2.0 specifications do not have support for identity-based primitives we can easily bootstrap identity-based private keys for Shamir’s signature scheme from regular RSA functionalities of TPM 2.0. In this way, key distribution becomes more secure as it is re-enforced by hardware and the rest of the functionalities can be carried from software implementations on mobile phones and in-vehicle controllers. We test the feasibility of the approach on modern Android devices and in-vehicle controllers as well as with a recent TPM circuit from Infineon.

We explore the use of delays to create a time-covert cryptographic authentication channel on the CAN bus. The use of clock skews has been recently proposed for detecting intrusions on CAN, using similar mechanisms that were previously exploited in computer or mobile networks in the past decade. However, the fine-grained control of timers easily allows controllers to adjust their clock potentially making such mechanisms ineffective as we argue here and was also proved by a recent research work. We exploit this potential shortcoming in a constructive sense, i.e., the accuracy of arrival times on in-vehicle buses and the fine-grained control of timer/counter circuits on automotive controllers allows us to use time as a covert channel to carry cryptographic authentication. Based on this procedure we propose an effective authentication and intrusion detection mechanism that is fully back-ward compatible with legacy implementations on CAN. Our proposal directly applies to any modern in-vehicle bus, e.g., CAN-FD, FlexRay, etc.

The 2018 FEIII Data Challenge aims to enhance a given knowledge graph by validating and enriching the set of competitor edges in the graph using multiple datasets. Upon an investigation of the data, we find that some of the competitor edges given as training data are inconsistent (e.g., conflicting with other relationships such as parent/subsidiary). Rather than using a machine learning approach that would have to address such difficulties and other ambiguities in the training data, we start by formulating two natural, semantic definitions of a competitor relationship. The first is a weak definition that is independent of the training data and identifies pairs of entities as potential competitors whenever in the same industry and geographical location, and provided that there is no negative evidence (such as the two entities being in the same family of companies). The second is a strong definition that intersects the pairs of entities obtained from the weak definition with the competitors given in the training dataset. These two definitions offer a framework implementation which can be extended to further utilize other attributes or additional information when available. One such extension that we can implement right away with the available data is to lift the competitor relationships from subsidiaries to their respective parent companies. We use a highlevel language (HIL) for entity linking to express and implement our two semantic definitions as well as the parent lifting extension. The resulting HIL algorithms are readable and easily extensible or modifiable by a domain expert. We show that our submission achieves 19.6% precision, 40.3% recall and 26.4% F1 score, and we make the case that with the availability of more data and more analytics these results can be further improved.

BACKGROUND AND AIM: Irritable bowel syndrome (IBS) has been associated with high prevalence of psychological and psychiatric disorders. However, the association between IBS and each of its subtypes ( diarrhea IBS-D, constipation IBS-C, mixed IBS-M) with anxiety, still remains unclear. The purpose of this study was to perform a comparative analysis, of the association between anxiety and IBS on a period of ten years.METHODS: PubMed was searched for studies analyzing IBS and anxiety, published at 10 years interval. The study presents a comparative analysis of the articles that were published between 2003-2005 and 2013-2015, investigating the correlation between anxiety and IBS.RESULTS: The initial search identified 220 articles, from which 156 were published between 2013 and 2015, and 64 were published between 2003 and 2005. Of these articles, 15 articles were included in the review. Out of these 15 articles, 10 articles analyzed the correlation between anxiety-depression status in IBS patients using specific questionnaires, 2 articles analyzed genetic variables in IBS, 1 article analyzed serom and monoamine oxidase levels in IBS, 1 article have analyzed seric levels of IL-1β and IL-10 in IBS, 1 article analyzed somatostatin and vasoactive intestinal peptide levels in IBS. The result was a review of 15 studies that analyzed the association between IBS and anxiety.CONCLUSIONS : IBS is a heterogeneous disorder caused by numerous psychological, immunological, infectious, endocrine and genetic factors. In recent years, the number of studies concentrating on genetic factors, cytokines and hormones has increased in comparison with the 2003-2005 period, when clinical investigation, using mainly questionnaires was the essential method. Also, the total number of papers investigating anxiety and IBS, considerably increased. The recent studies have confirmed the fact that IBS symptoms are often exacerbated during stressful events and the psychiatric treatment has a positive effect on gastro-intestinal symptomatology.

Entity resolution (ER) is the task of identifying different representations of the same real-world object across datasets. Designing and tuning ER algorithms is an error-prone, labor-intensive process, which can significantly benefit from data-driven, automated learning methods. Our focus is on "big data'' scenarios where the primary challenges include 1) identifying, out of a potentially massive set, a small subset of informative examples to be labeled by the user, 2) using the labeled examples to efficiently learn ER algorithms that achieve both high precision and high recall, and 3) executing the learned algorithm to determine duplicates at scale. Recent work on learning ER algorithms has employed active learning to partially address the above challenges by aiming to learn ER rules in the form of conjunctions of matching predicates, under precision guarantees. While successful in learning a single rule, prior work has been less successful in learning multiple rules that are sufficiently different from each other, thus missing opportunities for improving recall. In this paper, we introduce an active learning system that learns, at scale, multiple rules each having significant coverage of the space of duplicates, thus leading to high recall, in addition to high-precision. We show the superiority of our system on real-world ER scenarios of sizes up to tens of millions of records, over state-of-the-art active learning methods that learn either rules or committees of statistical classifiers for ER, and even over sophisticated methods based on first-order probabilistic models.

This paper is focused on two major problems within the ceramics industry. The reproduction of a desired color from pigments (which is time-intensive), and the correction of colors on the production line (which is costly) are processes which still rely heavily on numerous experiments carried out by human operators. This study presents the key aspects of these two processes and suggests some mathematical and computer sciences tools, aimed at automatizing the current procedures. Data was provided by an industrial partner, where the proposed models and solutions will be tested and validated. 2010 Mathematics Subject Classification: Primary 05A16, Secondary 05A15,05A18,11B75.

While cloud computing providers offer guaranteed allocations for resources such as CPU and memory, they do not offer any guarantees for network resources. The lack of network guarantees prevents tenants from predicting lower bounds on the performance of their applications. The research community has recognized this limitation but, unfortunately, prior solutions have significant limitations: either they are inefficient, because they are not work-conserving, or they are impractical, because they require expensive switch support or congestion-free network cores. In this paper, we propose ElasticSwitch, an efficient and practical approach for providing bandwidth guarantees. ElasticSwitch is efficient because it utilizes the spare bandwidth from unreserved capacity or underutilized reservations. ElasticSwitch is practical because it can be fully implemented in hypervisors, without requiring a specific topology or any support from switches. Because hypervisors operate mostly independently, there is no need for complex coordination between them or with a central controller. Our experiments, with a prototype implementation on a 100-server testbed, demonstrate that ElasticSwitch provides bandwidth guarantees and is work-conserving, even in challenging situations.

The performance of some Fourier transform based fundamental current detection algorithms is evaluated and compared in this paper. Some difficult conditions are imposed by a load which draws a highly distorted current with a slight fluctuation, i.e. with non-equal semi-periods. The implemented algorithms are Recursive Discrete Fourier Transform (RDFT), Discrete Fourier Transform (DFT) implemented through a Finite Impulse Response (FIR) filter and respectively the Fast Fourier Transform (FFT) with Butterfly operation and radix-2 decimation in time (DIT). A digital signal processor (DSP) based solution is used for real-time implementation, namely a dSPACE 1103 controller development board. MATLAB/Simulink software is used for the design of the algorithms because it provides specialized tools and blocks. For the performance comparison one used criteria like total harmonic distortion (THD) and settling time of detected fundamental current and also the run time for the DSP. This study may be useful as a guide for engineers in selecting the suitable Fourier transform based control algorithm for active power filters or other grid-synchronized equipments which are operating in difficult load conditions.

The paper deals with the behavior of a medium voltage/low voltage transformer (MV/LV), used to power the excitation circuit of a power group. The transformer's load is nonlinear, therefore inducing strong non-sinusoidal regimes in the primary and the secondary windings of the transformer. Non-sinusoidal regimes are due a thyristor rectifier used to power excitation. Two distinct situations are analyzed, corresponding for two contexts in which the three-phase transformer is operating. The recorded waveform from the transformer's primary and secondary windings, along with the harmonic decompositions of three-phase voltages and currents in the two windings are presented. The power quality factors from the two windings for both analyzed cases are presented. The transformer's losses are determined for both cases. Experimental determinations have shown that the use of cells for improving the power factor (in one of the analyzed cases) increased the harmonic distortion of voltages in both windings, also increasing the active power losses in them. When using the RC cells, the additional losses increased (through eddy currents). In normal operation (no RC cells), the global power factor is reduced both in primary and in secondary winding, and the current harmonics contribute in increasing the reactive power in the transformer's windings.

The paper deals with the components design and the simulation of a photovoltaic power generation system using MATLAB and Simulink software. The power plant is composed of photovoltaic panels connected in series and parallel strings, a DC-DC boost converter and a three-phase inverter which connects to a 0.4 kV three-phase low voltage grid and a 20 kV medium voltage grid by means of a step-up transformer. The DC-DC boost converter uses a MPPT controller and the inverter uses a control method based on d-q theory with a PI current regulator. Some cases for which the dynamic behavior of the configured photovoltaic system presents interest are simulated. They address the solar irradiance and temperature change.

We describe our approach to the FEIII Data Challenge, which requires matching entities across multiple financial datasets (FFIEC, SEC and LEI). By making use of a high-level language (HIL) that includes constructs for expressing both the matching logic and the policies to avoid or reduce the ambiguities among the matches, we are able to produce highly-accurate results in a sparse context, with only name and location attributes. As part of the high-level specification, we also make use of a Smart-Term Generation (STG) component, which provides us with a sophisticated subroutine for normalizing company names. The high-level specification is reusable, in the sense that the same HIL specification (modulo changing the attribute names) is uniformly applicable not only between FFIEC and SEC, but also between FFIEC and LEI, and between LEI and SEC.Our approach used only the data provided by the organizers, without analyzing any additional (external) datasets. For the task linking FFIEC records to SEC, we achieved 92.82% precision, 84.32% recall, and 88.38% F1-score. The precision and F1-score were the maximum reported across all participants, and recall was 1.3% less than the maximum 85.63%. For the task linking FFIEC records to LEI, we achieved 99.14% precision, 92.54% recall and 95.72% F1-score, with our F1-score 1.72% less than the maximum reported 97.44%. In this short paper, we provide a description of our method, together with an analysis of our results as well as possible directions for improvement.

Infrastructure-as-a-Service ("Cloud") data-centers intrinsically depend on high-performance networks to connect servers within the data-center and to the rest of the world. Cloud providers typically offer different service levels, and associated prices, for different sizes of virtual machine, memory, and disk storage. However, while all cloud providers provide network connectivity to tenant VMs, they seldom make any promises about network performance, and so cloud tenants suffer from highly-variable, unpredictable network performance. Many cloud customers do want to be able to rely on network performance guarantees, and many cloud providers would like to offer (and charge for) these guarantees. But nobody really agrees on how to define these guarantees, and it turns out to be challenging to define "network performance" in a way that is useful to both customers and providers. We attempt to bring some clarity to this question.

The paper analyzes the behavior of a three-phase consumer, consisting of several personal computers connected to a transformer's secondary phases. The modalities of monitoring and recording of electrical quantities are presented in this work. The structure of the consumer is described. The recorded waveforms are analyzed. The Fast Fourier Transform (FFT) analysis allowed for the decomposition of the electric signals - voltages and currents - in harmonic components, which are presented both graphically and tabular. The quality indices are computed, underlining the capacitive nature of the consumer, firstly on each phase, then globally. There are presented some conclusions related to the variation of the electric quantities due to the load variation. In order to estimate the states, Discrete Wavelet Transform (DWT) is used. Some indices related to the power quality for the acquired signals have been determined and are presented in a tabular form. The analyzed consumer introduced strong distorted and non-symmetrical regimes, affecting the network. This is leading to the conclusion that some intelligent recording systems need to be designed and realized, in order to correctly estimate the different states in real time.

This paper deals with issues related to a high power shunt active filter which is compensating the distorting regime and reactive power introduced by a silicon controlled rectifier. The current and voltage distortion is analyzed as a function of the system configuration and line reactors compensation. Different solutions are given for compensation and one of them is modeled and simulated in MATLAB/Simulink software environment. The proposed compensation system is composed by line reactors for the voltage notches introduced by the thyristor commutation, and a dual inverter shunt active power filter for the distortion regime and reactive power compensation. The advantages over the single inverter configuration are discussed. The most important one is that the harmonic current to be injected can be divided between the inverters, halving the power rating and lowering the high di/dt response demand of the PWM inductances. The used solver is continuous for an accurate representation of the plant but the active filter controller is acting like a digital one, being composed of MATLAB coded functions and discretized Simulink blocks at a fixed sample time, making possible the implementation in a real time hardware setup like dSPACE.

We present Midas, a system that uses complex data processing to extract and aggregate facts from a large collection of structured and unstructured documents into a set of unified, clean entities and relationships. Midas focuses on data for financial companies and is based on periodic filings with the U.S. Securities and Exchange Commission (SEC) and Federal Deposit Insurance Corporation (FDIC). We show that, by using data aggregated by Midas, we can provide valuable insights about financial institutions either at the whole system level or at the individual company level. The key technology components that we implemented in Midas and that enable the various financial applications are: information extraction, entity resolution, mapping and fusion, all on top of a scalable infrastructure based on Hadoop. We describe our experience in building the Midas system and also outline the key research questions that remain to be addressed towards building a generic, high-level infrastructure for large-scale data integration from public sources.

We introduce and develop a declarative framework for entity linking and, in particular, for entity resolution. As in some earlier approaches, our framework is based on a systematic use of constraints. However, the constraints we adopt are link-to-source constraints, unlike in earlier approaches where source-to-link constraints were used to dictate how to generate links. Our approach makes it possible to focus entirely on the intended properties of the outcome of entity linking, thus separating the constraints from any procedure of how to achieve that outcome. The core language consists of link-to-source constraints that specify the desired properties of a link relation in terms of source relations and built-in predicates such as similarity measures. A key feature of the link-to-source constraints is that they employ disjunction, which enables the declarative listing of all the reasons two entities should be linked. We also consider extensions of the core language that capture collective entity resolution by allowing interdependencies among the link relations. We identify a class of “good” solutions for entity-linking specifications, which we call maximum-value solutions and which capture the strength of a link by counting the reasons that justify it. We study natural algorithmic problems associated with these solutions, including the problem of enumerating the “good” solutions and the problem of finding the certain links, which are the links that appear in every “good” solution. We show that these problems are tractable for the core language but may become intractable once we allow interdependencies among the link relations. We also make some surprising connections between our declarative framework, which is deterministic, and probabilistic approaches such as ones based on Markov Logic Networks.

This paper presents measurements comprising both power quality and EMC fields for a static excitation system composed of a transformer and controlled rectifier bridge in a power system generator group. The replacement of the old auxiliary excitation generator eliminates the disadvantage of using electrical machines with the mass in motion inertia and wear over time, but with the price of facing two important problems: the line voltage notching due to thyristor commutation and also the line current distorting. FFT decomposition of the currents and voltages waveforms allows the obtaining of their harmonic spectrum in the range [0…2] kHz (PQ domain) and respectively [2…9] kHz (EMC domain). Theoretical aspects of line notching and some solutions for compensation strategies and their limitations are discussed.

There is a significant amount of "public" unstructured content available that is centered around the financial performance and counterparty relationships of a wide range of organizations, including private and public companies, governments, and public utilities. However, the wealth of structured entity and relationship information is buried inside a large amount of unstructured data. Converting such data into a structured format is essential for building novel analytics applications including counterparty and credit risk monitoring, investment decision making, development of new financial indices, systemic risk analysis, etc.In this paper, we illustrate three different application use cases where large-scale extraction and integration from the relevant public data enabled the creation of new applications to perform novel modeling and analysis. We discuss the main technological challenges encountered in such extraction and integration, and outline the common architectural platform used to build the three application use cases.

The real estate financial markets are complex supply chains. Understanding their behavior is limited by a lack of data that would capture the richly interconnected networks of financial institutions and complex financial products, e.g., asset backed securities. This lack of transparency is further compounded by limited knowledge of the contractual rules that control the flow of funds from mortgage pools to securities, as well as the financial events that regulate these flows. In this project, we will use the IBM Midas framework and tools to extract entities, relationships, events, contractual rules and risk profiles for financial institutions. Our source of information will be the MBS prospectus documents that are public and are filed with the Securities and Exchange Commission. We will describe the data management needs of the Haas Real Estate and Financial Markets (REFM) Lab and presents some recent REFM analytics that highlight the importance of these markets and the impact on systemic risk. We use excerpts extracted from the prospectus of a mortgage backed security (MBS) to illustrate the information extraction challenges and outline our approach to address these challenges.