Search

The intensity and extent of climate change impacts differ significantly with the geographical and ecological structure of the landscape. This is especially evident in mountain ecosystems where topographic, climatic and biological gradients make them extremely vulnerable to global environmental changes . Designing strategies to mitigate and adapt to global climate change on such local landscapes requires a context-specific vulnerabilities that take into account their particular characteristics. Presently, there are two main challenges in assessing climate change vulnerability in mountain ecosystems: 1) The models that are used for vulnerability assessments at global scales are being used at local scales with broad variables from few sectors that do not capture the range of characteristics of mountain ecosystems 2) indigenous knowledge about climate change are not considered in these models, which makes the implementation of mitigation/adaptation measures less successful. In this study, we highlight these issues drawing from our data collected in India’s Lesser Himalayan region (Darjeeling). We used a mixed research approach that combines a vulnerability assessment model with a participatory knowledge approach. We based climate change vulnerability around the socio-ecological system of the mountain landscape. The results from the interactive process showed that Darjeeling region is experiencing higher climate change vulnerability than the results produced by the model at the subregional level. We highlight critical variables that influence the socio-ecological system and need to be taken into account when assessing vulnerability and future adaptation scenarios. The study offers a decision support process for policymakers to plan climate mitigation/adaptation measures and future sustainability pathways.

Forecasting social media activity can be of practical use in many scenarios, from understanding trends, such as which topics are likely to engage more users in the coming week, to identifying unusual behavior, such as coordinated information operations or currency manipulation efforts. To evaluate a new approach to forecasting, it is important to have baselines against which to assess performance gains. We experimentally evaluate the performance of four baselines for forecasting activity in several social media datasets that record discussions related to three different geo-political contexts synchronously taking place on two different platforms, Twitter and YouTube. Experiments are done over hourly time periods. Our evaluation identifies the baselines which are most accurate for particular metrics and thus provides guidance for future work in social media modeling.

Wilkinson et al. claimed in previous work that Adherence of a dataset to the FAIR Guiding Principles enables its automated discovery. We present here a formalization of that claim, stating that all things of class “adherence to the FAIR Guiding principles” that are in the context of a thing of class “data set” can generally have a relation of type “enables” to a thing of class “automated discovery” in the same context.

DarkNet markets (DNMs) are flourishing, exploiting cryptocurrencies' pseudo-anonymity feature. To fight the illegal drug trade, it is necessary to understand the factors influencing choices made by DNM consumers. ClearNet forums facilitate interactions among new and experienced DarkNet users, thus serving as an ideal source for investigation and research. The methodology of temporal topic models and sentiment analysis were applied to the ClearNet forum data to explore DNM new drug users’ behavior over time. We inspected the development of major topics from the forum over eight years (2012–19) period to study the evolution in motivations and challenges faced by DNM users. The extracted temporal topic models successfully captured the apparent dynamics of critical events in the crypto world, showing the shift from Bitcoin to Monero. Our analysis demonstrates that the Bitcoin traceability announcement (made public in 2015) did not have a deterrent effect on the usage of Bitcoin, and this declaration instead acted as an advertisement for DNMs. Drug buyers used Bitcoin despite its traceability until Monero became untraceable in 2017 due to its privacy update.

The development of COVID-19 vaccines during the global pandemic that started in 2020 was marked by uncertainty and misinformation reflected also on social media. This paper provides a quantitative evaluation of the Uniform Resource Locators (URLs) shared on Twitter around the clinical trials of the AstraZeneca vaccine and their temporary interruption in September 2020. We analyzed URLs cited in Twitter messages before and after the temporary interruption of the vaccine development on September 9, 2020 to investigate the presence of low credibility and malicious information. We show that the halt of the AstraZeneca clinical trials prompted tweets that cast doubt, fear and vaccine opposition. We discovered a strong presence of URLs from low credibility or malicious websites, as classified by independent fact-checking organizations or identified by web hosting infrastructure features. Moreover, we identified what appears to be coordinated operations to artificially promote some of these URLs hosted on malicious websites.

Easy access to data is one of the main avenues to accelerate scientific research. As a key element of scientific innovations, data sharing allows the reproduction of results, helps prevent data fabrication, falsification, and misuse. Although the research benefits from data reuse are widely acknowledged, the data collections existing today are still kept in silos. Indeed, monitoring what happens to data once they have been handed to a third party is currently not feasible within the current data-sharing practices. We propose a blockchain-based system to trace data collections, and potentially create a more trustworthy data sharing process. In this paper, we present the LUCE (License accoUntability and CompliancE) architecture as a decentralized blockchain-based platform supporting data sharing and reuse. LUCE is designed to provide full transparency on what happens to the data after they are shared with third parties. The contributions of this work are: the definition of a generic model and an implementation for decentralized data sharing accountability and compliance and to incorporates dynamic consent and legal compliance mechanisms. We test the scalability of the platform in a real-time environment where a growing number of users access and reuse different datasets. Compared to existing data-sharing solutions, LUCE provides transparency over data sharing practices, enables data reuse and supports regulatory requirements. The experimentation shows that the platform can be scaled for a large number of users., 18 pages, 11 figures. arXiv admin note: text overlap with arXiv:1908.02287

Ever-increasing amount of data is generated by our citizens and used in our daily life every single day. These massive amounts of data can be used to improve digital technologies and develop data-driven innovations that can impact every aspect of people’s lives. However, lack of sharing of, access to and reuse of these data hinders the analysis possibilities and hence potential insights from the data. A number of challenges have been recognized such as technical barriers, security, data protection compliance to one or more legal jurisdictions, privacy concerns, and trust issues. This thesis aims to develop new privacy-preserving data sharing and analysis techniques that strengthen and extend the (re-)use of personal data while maximally protecting individuals’ privacy. To achieve this aim, this thesis addressed the research challenges on personal data sharing and use from perspectives of data organisations, scientific researchers, and individuals (citizens).

The mining of personal data collected by multiple organizations remains challenging in the presence of technical barriers, privacy concerns, and legal and/or organizational restrictions. While a number of privacy-preserving and data mining frameworks have recently emerged, much remains to show their practical utility. In this study, we implement and utilize a secure infrastructure using data from Statistics Netherlands and the Maastricht Study to learn the association between Type 2 Diabetes Mellitus (T2DM) and healthcare expenses considering the impact of lifestyle, physical activities, and complications of T2DM. Through experiments using real-world distributed personal data, we present the feasibility and effectiveness of the secure infrastructure for practical use cases of linking and analyzing vertically partitioned data across multiple organizations. We discovered that individuals diagnosed with T2DM had significantly higher expenses than those with prediabetes, while participants with prediabetes spent more than those without T2DM in all the included healthcare categories to different degrees. We further discuss a joint effort from technical, ethical-legal, and domain-specific experts that is highly valued for applying such a secure infrastructure to real-life use cases to protect data privacy.

In recent years, the work of organizations in the area of digitization has intensified significantly. This trend is also evident in the field of recruitment where job application tracking systems (ATS) have been developed to allow job advertisements to be published online. However, recent studies have shown that recruiting in most organizations is not inclusive, being subject to human biases and prejudices. Most discrimination activities appear early but subtly in the hiring process, for instance, exclusive phrasing in job advertisement discourages qualified applicants from minority groups from applying. The existing works are limited to analyzing, categorizing and highlighting the occurrence of bias in the recruitment process. In this paper, we go beyond this and develop machine learning models for identifying and classifying biased and discriminatory language in job descriptions. We develop and evaluate a machine learning system for identifying five major categories of biased and discriminatory language in job advertisements, i.e., masculine-coded, feminine-coded, exclusive, LGBTQ-coded, demographic and racial language. We utilized the combination of linguistic features with recent state-of-the-art word embeddings representations as input features for various machine learning classifiers. Our results show that the machine learning classifiers were able to identify all the five categories of biased and discriminatory language with a decent accuracy. The Random Forest classifier with FastText word embeddings achieved the best performance with tenfolds cross-validation. Our system directly addresses the bias in the attraction phase of hiring by identifying and classifying biased and discriminatory language and thus encouraging recruiters to write more inclusive job advertisements.

Linguists study language and language use in a range of settings and populations, yet they have not studied language, interaction, and communication behaviors and functions of the dying. This article argues that they should, using an account of the death of Gregory Bateson to make concrete the questions that could be asked, then showing some of the theoretical and practical contributions that the answers might make. The goal of such an endeavor would be to respectfully contribute a linguistic perspective to a core and truly universal human experience. (c) 2021 Published by Elsevier Ltd.

International audience; Taxonomies and ontologies for the characterization of everyday sounds have been developed in several research fields, including auditory cognition, soundscape research, artificial hearing, sound design, and medicine. Here, we surveyed 36 of such knowledge organization systems, which we identified through a systematic literature search. To evaluate the semantic domains covered by these systems within a homogeneous framework, we introduced a comprehensive set of verbal sound descriptors (sound source properties; attributes of sensation; sound signal descriptors; onomatopoeias; music genres), which we used to manually label the surveyed descriptor classes. We reveal that most taxonomies and ontologies were developed to characterize higher-level semantic relations between sound sources in terms of the sound-generating objects and actions involved (what/how), or in terms of the environmental context (where). This indicates the current lack of a comprehensive ontology of everyday sounds that covers simultaneously all semantic aspects of the relation between sounds. Such an ontology may have a wide range of applications and purposes, ranging from extending our scientific knowledge of auditory processes in the real world, to developing artificial hearing systems.

Today, content creators or influencers around the world are heavily reliant on digital platforms to mediate their revenue. Moreover, they use online platforms to gather funds for social causes through philanthropic campaigns. However, this raises concerns over the lack of transactional transparency of the campaigns and causes influencers to struggle attaining donors' trust and interest in charitable causes. In this paper, we present DePhi, a blockchain-based solution to ensure transparency in philanthropy campaigns. Using a blockchain network, we record donations in the form of micro-transactions paid to content creators through smart contracts. Moreover, we develop a browser extension through which donors can easily track their donations in real-time. We use the Interledger Protocol to develop payment options in different cryptocurrencies.

COVID-19 has affected more than 223 countries worldwide. There is a pressing need for non invasive, low costs and highly scalable solutions to detect COVID-19, especially in low-resource countries where PCR testing is not ubiquitously available. Our aim is to develop a deep learning model identifying COVID-19 using voice data recordings spontaneously provided by the general population (voice recordings and a short questionnaire) via their personal devices. The novelty of this work is in the development of a deep learning model for the identification of COVID-19 patients from voice recordings. Methods: We used the Cambridge University dataset consisting of 893 audio samples, crowd-sourced from 4352 participants that used a COVID-19 Sounds app. Voice features were extracted using a Mel-spectrogram analysis. Based on the voice data, we developed deep learning classification models to detect positive COVID-19 cases. These models included Long-Short Term Memory (LSTM) and Convolutional Neural Network (CNN). We compared their predictive power to baseline classification models, namely Logistic Regression and Support Vector Machine. Results: LSTM based on a Mel-frequency cepstral coefficients (MFCC) features achieved the highest accuracy (89%,) with a sensitivity and specificity of respectively 89% and 89%, The results achieved with the proposed model suggest a significant improvement in the prediction accuracy of COVID-19 diagnosis compared to the results obtained in the state of the art. Conclusion: Deep learning can detect subtle changes in the voice of COVID-19 patients with promising results. As an addition to the current testing techniques this model may aid health professionals in fast diagnosis and tracing of COVID-19 cases using simple voice analysis, Comment: 9 pages, 6 figures, poster presented at the European Respiratory Society, Barcelona, September 2022

Modeling social media activity has numerous practical implications such as in helping analyze strategic information operations, designing intervention techniques to mitigate disinformation, or delivering critical information during disaster relief operations. In this paper, we propose a modeling technique that forecasts topic-specific daily volume of social media activities by multiplexing different exogenous signals, such as news reports and armed conflicts records, and endogenous data from the social media platform we model. For this, we trained a collection of LSTM models, each leveraging a different exogenous source, and dynamically select one model for each topic. Empirical evaluations with real datasets from two social media platforms and two different contexts each composed of multiple interrelated topics demonstrate the effectiveness of our solution.

PurposeA vast body of literature has documented the negative consequences of stress on employee performance and well-being. These deleterious effects are particularly pronounced for service agents who need to constantly endure and manage customer emotions. The purpose of this paper is to introduce and describe a deep learning model to predict in real-time service agent stress from emotion patterns in voice-to-voice service interactions.Design/methodology/approachA deep learning model was developed to identify emotion patterns in call center interactions based on 363 recorded service interactions, subdivided in 27,889 manually expert-labeled three-second audio snippets. In a second step, the deep learning model was deployed in a call center for a period of one month to be further trained by the data collected from 40 service agents in another 4,672 service interactions.FindingsThe deep learning emotion classifier reached a balanced accuracy of 68% in predicting discrete emotions in service interactions. Integrating this model in a binary classification model, it was able to predict service agent stress with a balanced accuracy of 80%.Practical implicationsService managers can benefit from employing the deep learning model to continuously and unobtrusively monitor the stress level of their service agents with numerous practical applications, including real-time early warning systems for service agents, customized training and automatically linking stress to customer-related outcomes.Originality/valueThe present study is the first to document an artificial intelligence (AI)-based model that is able to identify emotions in natural (i.e. nonstaged) interactions. It is further a pioneer in developing a smart emotion-based stress measure for service agents. Finally, the study contributes to the literature on the role of emotions in service interactions and employee stress.

In the course of under a year, the European Commission has launched some of the most important regulatory proposals to date on platform governance. The Commission's goals behind cross-sectoral regulation of this sort include the protection of markets and democracies alike. While all these acts propose sophisticated rules for setting up new enforcement institutions and procedures, one aspect remains highly unclear: how digital enforcement will actually take place in practice. Focusing on the Digital Services Act (DSA), this discussion paper critically addresses issues around social media data access for the purpose of digital enforcement and proposes the use of a legal compliance application programming interface (API) as a means to facilitate compliance with the DSA and complementary European and national regulation. To contextualize this discussion, the paper pursues two scenarios that exemplify the harms arising out of content monetization affecting a particularly vulnerable category of social media users: children. The two scenarios are used to further reflect upon essential issues surrounding data access and legal compliance with the DSA and further applicable legal standards in the field of labour and consumer law., Accepted for publication at ACM FAccT Conference 2022

Purpose Magnetic Resonance Imaging (MRI) provides an essential contribution in the screening, detection, diagnosis, staging, treatment and follow-up in patients with a neurological neoplasm. Deep learning (DL), a subdomain of artificial intelligence has the potential to enhance the characterization, processing and interpretation of MRI images. The aim of this review paper is to give an overview of the current state-of-art usage of DL in MRI for neuro-oncology. Methods We reviewed the Pubmed database by applying a specific search strategy including the combination of MRI, DL, neuro-oncology and its corresponding search terminologies, by focussing on Medical Subject Headings (Mesh) or title/abstract appearance. The original research papers were classified based on its application, into three categories: technological innovation, diagnosis and follow-up. Results Forty-one publications were eligible for review, all were published after the year 2016. The majority (N = 22) was assigned to technological innovation, twelve had a focus on diagnosis and seven were related to patient follow-up. Applications ranged from improving the acquisition, synthetic CT generation, auto-segmentation, tumor classification, outcome prediction and response assessment. The majority of publications made use of standard (T1w, cT1w, T2w and FLAIR imaging), with only a few exceptions using more advanced MRI technologies. The majority of studies used a variation on convolution neural network (CNN) architectures. Conclusion Deep learning in MRI for neuro-oncology is a novel field of research; it has potential in a broad range of applications. Remaining challenges include the accessibility of large imaging datasets, the applicability across institutes/vendors and the validation and implementation of these technologies in clinical practise.

The quality of a Knowledge Graph (also known as Linked Data) is an important aspect to indicate its fitness for use in an application. Several quality dimensions are identified, such as accuracy, completeness, timeliness, provenance, and accessibility, which are used to assess the quality. While many prior studies offer a landscape view of data quality dimensions, here we focus on presenting a systematic literature review for assessing the completeness of Knowledge Graph. We gather existing approaches from the literature and analyze them qualitatively and quantitatively. In particular, we unify and formalize commonly used terminologies across 56 articles related to the completeness dimension of data quality and provide a comprehensive list of methodologies and metrics used to evaluate the different types of completeness. We identify seven types of completeness, including three types that were not previously identified in previous surveys. We also analyze nine different tools capable of assessing Knowledge Graph completeness. The aim of this Systematic Literature Review is to provide researchers and data curators a comprehensive and deeper understanding of existing works on completeness and its properties, thereby encouraging further experimentation and development of new approaches focused on completeness as a data quality dimension of Knowledge Graph.

This short discussion paper addresses how controversy is monetized online by reflecting on a new iteration of the shock value in media production, identified on social media as the ‘clout chasing’ phenomenon. We first exemplify controversial behavior, and subsequently proceed to defining clout chasing, which we discuss this concept in relation to existing frameworks for the understanding of controversy on social media. We then outline what clout chasing entails as a content monetization strategy, and address the risks associated with this approach. In doing so, we introduce the concept of ‘content self-moderation,’ which encompasses how creators use content moderation as a way to hedge monetization risks arising out of their reliance on controversy for economic growth. This concept is discussed in the context of the automated content governance entailed by algorithmic platform architectures, to contribute to existing scholarship on platform governance.

The past 25 years have seen many attempts to introduce defeasible-reasoning capabilities into a description logic setting. Many, if not most, of these attempts are based on preferential extensions of description logics, with a significant number of these, in turn, following the so-called KLM approach to defeasible reasoning initially advocated for propositional logic by Kraus, Lehmann, and Magidor. Each of these attempts has its own aim of investigating particular constructions and variants of the (KLM-style) preferential approach. Here our aim is to provide a comprehensive study of the formal foundations of preferential defeasible reasoning for description logics in the KLM tradition. We start by investigating a notion of defeasible subsumption in the spirit of defeasible conditionals as studied by Kraus, Lehmann, and Magidor in the propositional case. In particular, we consider a natural and intuitive semantics for defeasible subsumption, and we investigate KLM-style syntactic properties for both preferential and rational subsumption. Our contribution includes two representation results linking our semantic constructions to the set of preferential and rational properties considered. Besides showing that our semantics is appropriate, these results pave the way for more effective decision procedures for defeasible reasoning in description logics. Indeed, we also analyse the problem of non-monotonic reasoning in description logics at the level of entailment and present an algorithm for the computation of rational closure of a defeasible knowledge base. Importantly, our algorithm relies completely on classical entailment and shows that the computational complexity of reasoning over defeasible knowledge bases is no worse than that of reasoning in the underlying classical DL ALC .

Background The European Platform on Rare Disease Registration (EU RD Platform) aims to address the fragmentation of European rare disease (RD) patient data, scattered among hundreds of independent and non-coordinating registries, by establishing standards for integration and interoperability. The first practical output of this effort was a set of 16 Common Data Elements (CDEs) that should be implemented by all RD registries. Interoperability, however, requires decisions beyond data elements - including data models, formats, and semantics. Within the European Joint Programme on Rare Diseases (EJP RD), we aim to further the goals of the EU RD Platform by generating reusable RD semantic model templates that follow the FAIR Data Principles. Results Through a team-based iterative approach, we created semantically grounded models to represent each of the CDEs, using the SemanticScience Integrated Ontology as the core framework for representing the entities and their relationships. Within that framework, we mapped the concepts represented in the CDEs, and their possible values, into domain ontologies such as the Orphanet Rare Disease Ontology, Human Phenotype Ontology and National Cancer Institute Thesaurus. Finally, we created an exemplar, reusable ETL pipeline that we will be deploying over these non-coordinating data repositories to assist them in creating model-compliant FAIR data without requiring site-specific coding nor expertise in Linked Data or FAIR. Conclusions Within the EJP RD project, we determined that creating reusable, expert-designed templates reduced or eliminated the requirement for our participating biomedical domain experts and rare disease data hosts to understand OWL semantics. This enabled them to publish highly expressive FAIR data using tools and approaches that were already familiar to them.

BackgroundSign language is an essential means of communication for hearing-impaired individuals.ObjectiveWe aimed to develop an American sign language recognition dataset and use it in the deep learning model which depends on neural networks to interpret gestures of sign language and hand poses to natural language.MethodsIn this study, we developed a dataset and a Convolutional Neural Network-based sign language interface system to interpret gestures of sign language and hand poses to natural language. The neural network developed in this study is a Convolutional Neural Network (CNN) which enhances the predictability of the American Sign Language alphabet (ASLA). This research establishes a new dataset of the American Sign Language alphabet which takes into consideration various conditions such as lighting and distance.ResultsThe dataset created in this study is a new addition in the field of sign language recognition (SLR). This dataset may be used to develop SLR systems. Furthermore, our research compares the results of our dataset with two different datasets from other studies. The other datasets have invariant scene conditions, but our suggested CNN model demonstrated high accuracy for all the tested datasets. Despite the different conditions and volume of the new dataset, it achieved 99.38% accuracy with excellent prediction and small loss (0.0250).ConclusionsThe proposed system may be considered a promising solution in medical applications that use deep learning with superior accuracy. Moreover, our dataset was created under variable conditions which increases the number of contributions, comparisons, results and conclusions in the field of SLR and may enhance such systems.

The work presented in this Chapter was designed for the purposes of a specific European project, named “CrossJustice” (hereinafter cj). The first part of the Chapter focuses on the modelling of definitions within the six procedural European Union Directives, which were the focus of the research project. The goal is that of illustrating the steps and the challenges towards the construction of a lightweight ontology representing this information. The process of building the ontology revealed a number of aspects that present opportunities for further research. In the second part, an automatic experimentation concerning the harmonization of EU directives is presented, built on top of an automatic analysis of national implementations for each pair of states to investigate the similarity of the corresponding texts. The applied text mining and natural language processing techniques then culminate in the computation of the cosine similarity between vectors associated with legal texts. On that basis, an aggregated index approximates the degree of harmonization within the six EU Directives.

Introduction: The course of OCD differs widely among OCD patients, varying from chronic symptoms to full remission. No tools for individual prediction of OCD remission are currently available. This study aimed to develop a machine learning algorithm to predict OCD remission after two years, using solely predictors easily accessible in the daily clinical routine. Methods: Subjects were recruited in a longitudinal multi-center study (NOCDA). Gradient boosted decision trees were used as supervised machine learning technique. The training of the algorithm was performed with 227 predictors and 213 observations collected in a single clinical center. Hyper-parameter optimization was performed with cross-validation and a Bayesian optimization strategy. The predictive performance of the algorithm was subsequently tested in an independent sample of 215 observations collected in five different centers. Between-center differences were investigated with a bootstrap resampling approach. Results: The average predictive performance of the algorithm in the test centers resulted in an AUROC of 0.7820, a sensitivity of 73.42%, and a specificity of 71.45%. Results also showed a significant between-center variation in the predictive performance. The most important predictors resulted related to OCD severity, OCD chronic course, use of psychotropic medications, and better global functioning. Limitations: All recruiting centers followed the same assessment protocol and are in The Netherlands. Moreover, the sample of the data recruited in some of the test centers was limited in size. Discussion: The algorithm demonstrated a moderate average predictive performance, and future studies will focus on increasing the stability of the predictive performance across clinical settings.

Many data sets include categorical features in the form of nominal and ordinal features. However, most machine learning algorithms cannot deal with categorical features directly because they require numerical input features. Categorical embeddings are an effective approach to converting categorical features into numerical vectors. This work proposes a novel embedding approach, called Semantic Correlation Graph Embedding, to create embeddings from knowledge graphs. The approach constructs a semantic correlation graph of triplets among the categorical features to learn numerical embeddings. Our approach aims to uncover relationships taking place in categorical data in terms of low-level knowledge and semantics that may help group the features of the data sets under semantic entities. Three distinct embedding models are proposed according to how the graph is constructed. The results are evaluated with two public data sets. They show that the learned embeddings produce a statistically significant improvement in the performance of the classification tasks in terms of AUC, F1 score, precision, and recall.

This work proposes a heterophily-based metric for quantifying polarization in social networks where multiple ideological, antagonistic communities coexist. This metric captures node-level polarization and is built on user’s affinity towards other communities rather than their own. Node-level values can then be aggregated at the community, network, or sub-network level, providing a more detailed map of polarization. We tested our metric on the Polblogs network, White Helmets Twitter interaction network with two communities and the VoterFraud2020 domain network with five communities. We also tested our metric on dK-random graphs to verify that it results in low polarization scores, as expected. Finally, we compared our metric with two widely used polarization measures: Guerra’s polarization index and RWC.

Recently a huge number of knowledge graphs (KGs) has been generated, but there has not been enough attention to generate high-quality metadata to enable users to reuse the KGs for their own purposes. The main challenge is to generate standardized and high quality descriptive metadata which helps users understand the content of the large KGs. Some existing solutions make use of a combination of schema-level patterns derived from graph summarization with instance-level snippets. I will follow this trend and develop a method based on a combination of content-based patterns with user activity data such as SPARQL query logs to make generated metadata more informative and useful than other developed approaches. The problem of current models is generating complex, long or insufficient metadata which I plan to tackle by proposing a guideline to generate standard metadata during my Ph.D.

The popularity of location-based social networks (LBSNs) has led to a tremendous amount of user check-in data. Recommending points of interest (POIs) plays a key role in satisfying users' needs in LBSNs. While recent work has explored the idea of adopting collaborative ranking (CR) for recommendation, there have been few attempts to incorporate temporal information for POI recommendation using CR. In this article, we propose a two-phase CR algorithm that incorporates the geographical influence of POIs and is regularized based on the variance of POIs popularity and users' activities over time. The time-sensitive regularizer penalizes user and POIs that have been more time-sensitive in the past, helping the model to account for their long-term behavioral patterns while learning from user-POI interactions. Moreover, in the first phase, it attempts to rank visited POIs higher than the unvisited ones, and at the same time, apply the geographical influence. In the second phase, our algorithm tries to rank users' favorite POIs higher on the recommendation list. Both phases employ a collaborative learning strategy that enables the model to capture complex latent associations from two different perspectives. Experiments on real-world datasets show that our proposed time-sensitive collaborative ranking model beats state-of-the-art POI recommendation methods., To appear in IEEE Transactions on Knowledge and Data Engineering (TKDE)

In dichotomous item response theory (IRT) framework, the asymptotic standard error (ASE) is the most common statistic to evaluate the precision of various ability estimators. Easy-to-use ASE formulas are readily available; however, the accuracy of some of these formulas was recently questioned and new ASE formulas were derived from a general asymptotic theory framework. Furthermore, exact standard errors were suggested to better evaluate the precision of ability estimators, especially with short tests for which the asymptotic framework is invalid. Unfortunately, the accuracy of exact standard errors was assessed so far only in a very limiting setting. The purpose of this article is to perform a global comparison of exact versus (classical and new formulations of) asymptotic standard errors, for a wide range of usual IRT ability estimators, IRT models, and with short tests. Results indicate that exact standard errors globally outperform the ASE versions in terms of reduced bias and root mean square error, while the new ASE formulas are also globally less biased than their classical counterparts. Further discussion about the usefulness and practical computation of exact standard errors are outlined.

This editorial summarizes the content of the Special Issue on Quality Assessment of Knowledge Graphs of the Journal of Data and Information Quality (JDIQ). We dedicate this special issue to the memory of our colleague and friend Amrapali Zaveri.

PurposeWith the advent of increasingly sophisticated AI, the nature of work in the service frontline is changing. The next frontier is to go beyond replacing routine tasks and augmenting service employees with AI. The purpose of this paper is to investigate whether service employees augmented with AI-based emotion recognition software are more effective in interpersonal emotion regulation (IER) and whether and how IER impacts their own affective well-being.Design/methodology/approachFor the underlying study, an AI-based emotion recognition software was developed in order to assist service employees in managing customer emotions. A field study based on 2,459 call center service interactions assessed the effectiveness of the AI in augmenting service employees for IER and the immediate downstream consequences for well-being relevant outcomes.FindingsAugmenting service employees with AI significantly improved their IER activities. Employees in the AI (vs control) condition were significantly more effective in regulating customer emotions. IER goal attainment, in turn, mediated the effect on employee affective well-being. Perceived stress related to exposure to the AI augmentation acted as a competing mediator.Practical implicationsService firms can benefit from state-of-the-art AI technology by focusing on its capacity to augment rather than merely replacing employees. Furthermore, signaling IER goal attainment with the help of technology may provide uplifting consequences for service employee affective well-being.Originality/valueThe present study is among the first to empirically test the introduction of an AI-fueled technology to augment service employees in handling customer emotions. This paper further complements the literature by investigating IER in a real-life setting and by uncovering goal attainment as a new mechanism underlying the effect of IER on the well-being of the sender.

Highlights • Nitroglycerin didn’t improve overall survival of NSCLC patients. • The toxicity of combining nitroglycerin with standard treatment was mild. • Increased uptake of HX4 showed negative prognostic significance in NSCLC patients. • Tumor perfusion after nitroglycerin treatment did not correlate with outcome., Background Nitroglycerin is proposed as an agent to reduce tumour hypoxia by improving tumour perfusion. We investigated the potential of nitroglycerin as a radio-sensitizer in non-small cell lung cancer (NSCLC) and the potential of functional imaging for patient selection. Material and methods Trial NCT01210378 is a single arm phase II trial, designed to detect 15% improvement in 2-year overall survival (primary endpoint) in stage IB-IV NSCLC patients treated with radical (chemo-) radiotherapy and a Transiderm-Nitro 5 patch during radiotherapy. Patients underwent dynamic contrast-enhanced CTs (DCE-CT) and HX4 (hypoxia) PET/CTs before and after nitroglycerin. Secondary endpoints were progression-free survival, toxicity and the prognostic value of tumour perfusion/hypoxia at baseline and after nitroglycerin. Results The trial stopped after a futility analysis after 42 patients. At median follow-up of 41 months, two-year and median OS were 58% (95% CI: 44–78%) and 38 months (95% CI: 22–54 months), respectively. Nitroglycerin could not reduce tumour hypoxia. DCE-CT parameters did not correlate with OS, whereas hypoxic tumours had a worse OS (p = 0.029). Changes in high-uptake fraction of HX4 and tumour blood flow were negatively correlated (r = -0.650, p = 0.022). The heterogeneity in treatment modalities and patient characteristics combined with a small sample size made further subgroup analysis of survival results impossible. Toxicity related to nitroglyerin was limited to headache (17%) and hypotension (2.4%). Conclusion Nitroglycerin did not improve OS of NSCLC patients treated with (chemo-)radiotherapy. A general ability of nitroglycerin to reduce hypoxia was not shown.

We investigated whether genetic susceptibility to tuberculosis (TB) influences lung adenocarcinoma development among never-smokers using TB genome-wide association study (GWAS) results within the Female Lung Cancer Consortium in Asia. Pathway analysis with the adaptive rank truncated product method was used to assess the association between a TB-related gene-set and lung adenocarcinoma using GWAS data from 5512 lung adenocarcinoma cases and 6277 controls. The gene-set consisted of 31 genes containing known/suggestive associations with genetic variants from previous TB-GWAS. Subsequently, we followed-up with Mendelian Randomization to evaluate the association between TB and lung adenocarcinoma using three genome-wide significant variants from previous TB-GWAS in East Asians. The TB-related gene-set was associated with lung adenocarcinoma (p = 0.016). Additionally, the Mendelian Randomization showed an association between TB and lung adenocarcinoma (OR = 1.31, 95% CI: 1.03, 1.66, p = 0.027). Our findings support TB as a causal risk factor for lung cancer development among never-smoking Asian women.

The FAIR principles were received with broad acceptance in several scientific communities. However, there is still some degree of uncertainty on how they should be implemented. Several self-report questionnaires have been proposed to assess the implementation of the FAIR principles. Moreover, the FAIRmetrics group released 14, general-purpose maturity for representing FAIRness. Initially, these metrics were conducted as open-answer questionnaires. Recently, these metrics have been implemented into a software that can automatically harvest metadata from metadata providers and generate a principle-specific FAIRness evaluation. With so many different approaches for FAIRness evaluations, we believe that further clarification on their limitations and advantages, as well as on their interpretation and interplay should be considered.

Background and purpose: Access to healthcare data is indispensable for scientific progress and innovation. Sharing healthcare data is time-consuming and notoriously difficult due to privacy and regulatory concerns. The Personal Health Train (PHT) provides a privacy-by-design infrastructure connecting FAIR (Findable, Accessible, Interoperable, Reusable) data sources and allows distributed data analysis and machine learning. Patient data never leaves a healthcare institute.Materials and methods: Lung cancer patient-specific databases (tumor staging and post-treatment survival information) of oncology departments were translated according to a FAIR data model and stored locally in a graph database. Software was installed locally to enable deployment of distributed machine learning algorithms via a central server. Algorithms (MATLAB, code and documentation publicly available) are patient privacy-preserving as only summary statistics and regression coefficients are exchanged with the central server. A logistic regression model to predict post-treatment two-year survival was trained and evaluated by receiver operating characteristic curves (ROC), root mean square prediction error (RMSE) and calibration plots.Results: In 4 months, we connected databases with 23 203 patient cases across 8 healthcare institutes in 5 countries (Amsterdam, Cardiff, Maastricht, Manchester, Nijmegen, Rome, Rotterdam, Shanghai) using the PHT. Summary statistics were computed across databases. A distributed logistic regression model predicting post-treatment two-year survival was trained on 14 810 patients treated between 1978 and 2011 and validated on 8 393 patients treated between 2012 and 2015.Conclusion: The PHT infrastructure demonstrably overcomes patient privacy barriers to healthcare data sharing and enables fast data analyses across multiple institutes from different countries with different regulatory regimens. This infrastructure promotes global evidence-based medicine while prioritizing patient privacy. (C) 2019 The Authors. Published by Elsevier B.V.

Contains fulltext : 229874.pdf (Publisher’s version ) (Open Access) PURPOSE: One of the most frequently cited radiomics investigations showed that features automatically extracted from routine clinical images could be used in prognostic modeling. These images have been made publicly accessible via The Cancer Imaging Archive (TCIA). There have been numerous requests for additional explanatory metadata on the following datasets - RIDER, Interobserver, Lung1, and Head-Neck1. To support repeatability, reproducibility, generalizability, and transparency in radiomics research, we publish the subjects' clinical data, extracted radiomics features, and digital imaging and communications in medicine (DICOM) headers of these four datasets with descriptive metadata, in order to be more compliant with findable, accessible, interoperable, and reusable (FAIR) data management principles. ACQUISITION AND VALIDATION METHODS: Overall survival time intervals were updated using a national citizens registry after internal ethics board approval. Spatial offsets of the primary gross tumor volume (GTV) regions of interest (ROIs) associated with the Lung1 CT series were improved on the TCIA. GTV radiomics features were extracted using the open-source Ontology-Guided Radiomics Analysis Workflow (O-RAW). We reshaped the output of O-RAW to map features and extraction settings to the latest version of Radiomics Ontology, so as to be consistent with the Image Biomarker Standardization Initiative (IBSI). Digital imaging and communications in medicine metadata was extracted using a research version of Semantic DICOM (SOHARD, GmbH, Fuerth; Germany). Subjects' clinical data were described with metadata using the Radiation Oncology Ontology. All of the above were published in Resource Descriptor Format (RDF), that is, triples. Example SPARQL queries are shared with the reader to use on the online triples archive, which are intended to illustrate how to exploit this data submission. DATA FORMAT: The accumulated RDF data are publicly accessible through a SPARQL endpoint where the triples are archived. The endpoint is remotely queried through a graph database web application at http://sparql.cancerdata.org. SPARQL queries are intrinsically federated, such that we can efficiently cross-reference clinical, DICOM, and radiomics data within a single query, while being agnostic to the original data format and coding system. The federated queries work in the same way even if the RDF data were partitioned across multiple servers and dispersed physical locations. POTENTIAL APPLICATIONS: The public availability of these data resources is intended to support radiomics features replication, repeatability, and reproducibility studies by the academic community. The example SPARQL queries may be freely used and modified by readers depending on their research question. Data interoperability and reusability are supported by referencing existing public ontologies. The RDF data are readily findable and accessible through the aforementioned link. Scripts used to create the RDF are made available at a code repository linked to this submission: https://gitlab.com/UM-CDS/FAIR-compliant_clinical_radiomics_and_DICOM_metadata.

In modern healthcare systems, the ability to share electronic health records is crucial for providing quality care and for enabling a larger spectrum of health services. Health data sharing is dependent on obtaining individual consent which, in turn, is hindered by a lack of resources. To this extent, blockchain-based platforms facilitate data sharing by creating a trusted distributed network of users. These users are enabled to share their data without depending on the time and resources of specific actors (such as the health services). In blockchain-based platforms, data governance mechanisms become very important due to the need to specify and monitor data sharing and data use conditions. In this article, we present a blockchain-based data-sharing consent model for access control over individual health data. We use smart contracts to dynamically represent the individual's consent over health data and to enable data requesters to search and access those data. The dynamic consent model extends to two ontologies: the Data Use Ontology (DUO) which models the individual consent of users and the Automatable Discovery and Access Matrix (ADA-M), which describes queries from data requesters. We deploy the model on Ethereum blockchain and evaluate different data sharing scenarios. The contribution of this article consists of the creation of an individual consent model for health data sharing platforms. Such a model guarantees that individual consent is respected and that all the participants in the data sharing platform are accountable. The evaluation of our solution indicates that such a data sharing model provides a flexible decision-making approach over data usage by data requesters. Our experimental evaluation shows that the proposed model is efficient and adapts to personalized access control policies in different data-sharing scenarios.

This paper focuses on fine-grained, secure access to FAIR data, for which we propose ontology-based data access policies. These policies take into account both the FAIR aspects of the data relevant to access (such as provenance and licence), expressed as metadata, and additional metadata describing users. With this tripartite approach (data, associated metadata expressing FAIR information, and additional metadata about users), secure and controlled access to object data can be obtained. This yields a security dimension to the “A” (accessible) in FAIR, which is clearly needed in domains like security and intelligence. These domains need data to be shared under tight controls, with widely varying individual access rights. In this paper, we propose an approach called Ontology-Based Access Control (OBAC), which utilizes concepts and relations from a data set's domain ontology. We argue that ontology-based access policies contribute to data reusability and can be reconciled with privacy-aware data access policies. We illustrate our OBAC approach through a proof-of-concept and propose that OBAC to be adopted as a best practice for access management of FAIR data.

This contribution explores the application of data science and artificial intelligence to legal research, more specifically an element that has not received much attention: the research infrastructure required to make such analysis possible. In recent years, EU law has become increasingly digitised and published in online databases such as EUR-Lex and HUDOC. However, the main barrier inhibiting legal scholars from analysing this information is lack of training in data analytics. Legal analytics software can mitigate this problem to an extent. However, current systems are dominated by the commercial sector. In addition, most systems focus on search of legal information but do not facilitate advanced visualisation and analytics. Finally, free to use systems that do provide such features are either too complex to use for general legal scholars, or are not rich enough in their analytics tools. In this paper, we motivate the case for building a software platform that addresses these limitations. Such software can provide a powerful platform for visualising and exploring connections and correlations in EU case law, helping to unravel the “DNA” behind EU legal systems. It will also serve to train researchers and students in schools and universities to analyse legal information using state-of-the-art methods in data science, without requiring technical proficiency in the underlying methods. We also suggest that the software should be powered by a data infrastructure and management paradigm following the seminal FAIR (Findable, Accessible, Interoperable and Reusable) principles.

Atlas-based automatic segmentation is used in radiotherapy planning to accelerate the delineation of organs at risk (OARs). Atlas selection has been proposed as a way to improve the accuracy and execution time of segmentation, assuming that, the more similar the atlas is to the patient, the better the results will be. This paper presents an analysis of atlas selection methods in the context of radiotherapy treatment planning. For a range of commonly contoured OARs, a thorough comparison of a large class of typical atlas selection methods has been performed. For this evaluation, clinically contoured CT images of the head and neck ( ${N}={316}$ ) and thorax ( ${N}={280}$ ) were used. The state-of-the-art intensity and deformation similarity-based atlas selection methods were found to compare poorly to perfect atlas selection. Counter-intuitively, atlas selection methods based on a fixed set of representative atlases outperformed atlas selection methods based on the patient image. This study suggests that atlas-based segmentation with currently available selection methods compares poorly to the potential best performance, hampering the clinical utility of atlas-based segmentation. Effective atlas selection remains an open challenge in atlas-based segmentation for radiotherapy planning.

While there exists an abundance of open biomedical data, the lack of high-quality metadata makes it challenging for others to find relevant datasets and to reuse them for another purpose. In particular, metadata are useful to understand the nature and provenance of the data. A common approach to improving the quality of metadata relies on expensive human curation, which itself is time-consuming and also prone to error. Towards improving the quality of metadata, we use scientific publications to automatically predict metadata key:value pairs. For prediction, we use a Convolutional Neural Network (CNN) and a Bidirectional Long-short term memory network (BiLSTM). We focus our attention on the NCBI Disease Corpus, which is used for training the CNN and BiLSTM. We perform two different kinds of experiments with these two architectures: (1) we predict the disease names by using their unique ID in the MeSH ontology and (2) we use the tree structures of MeSH ontology to move up in the hierarchy of these disease terms, which reduces the number of labels. We also perform various multi-label classification techniques for the above-mentioned experiments. We find that in both cases CNN achieves the best results in predicting the superclasses for disease with an accuracy of 83%.

ABSTR A C T The effectiveness of machine learning models to provide accurate and consistent results in drug discovery and clinical decision support is strongly dependent on the quality of the data used. However, substantive amounts of open data that drive drug discovery suffer from a number of issues including inconsistent representation, inaccurate reporting, and incomplete context. For example, databases of FDA-approved drug indications used in computational drug repositioning studies do not distinguish between treatments that simply offer symptomatic relief from those that target the underlying pathology. Moreover, drug indication sources often lack proper provenance and have little overlap. Consequently, new predictions can be of poor quality as they offer little in the way of new insights. Hence, work remains to be done to establish higher quality databases of drug indications that are suitable for use in drug discovery and repositioning studies. Here, we report on the combination of weak supervision (i.e., programmatic labeling and crowdsourcing) and deep learning methods for relation extraction from DailyMed text to create a higher quality drug-disease relation dataset. The generated drug-disease relation data shows a high overlap with DrugCentral, a manually curated dataset. Using this dataset, we constructed a machine learning model to classify relations between drugs and diseases from text into four categories; treat-ment, symptomatic relief, contradiction, and effect, exhibiting an improvement of 15.5% with Bi-LSTM (F1 score of 71.8%) over the best performing discrete method. Access to high quality data is crucial to building accurate and reliable drug repurposing prediction models. Our work suggests how the combination of crowds, experts, and machine learning methods can go hand-in-hand to improve datasets and predictive models.

Transparent evaluations of FAIRness are increasingly required by a wide range of stakeholders, from scientists to publishers, funding agencies and policy makers. We propose a scalable, automatable framework to evaluate digital resources that encompasses measurable indicators, open source tools, and participation guidelines, which come together to accommodate domain relevant community-defined FAIR assessments. The components of the framework are: (1) Maturity Indicators - community-authored specifications that delimit a specific automatically-measurable FAIR behavior; (2) Compliance Tests - small Web apps that test digital resources against individual Maturity Indicators; and (3) the Evaluator, a Web application that registers, assembles, and applies community-relevant sets of Compliance Tests against a digital resource, and provides a detailed report about what a machine “sees” when it visits that resource. We discuss the technical and social considerations of FAIR assessments, and how this translates to our community-driven infrastructure. We then illustrate how the output of the Evaluator tool can serve as a roadmap to assist data stewards to incrementally and realistically improve the FAIRness of their resources.

It has become increasingly easy to collect data from individuals over long periods of time. Examples include smart-phone applications used to track movements with GPS, web-log data tracking individuals' browsing behavior, and longitudinal (cohort) studies where many individuals are monitored over an extensive period of time. All these datasets cover a large number of individuals and collect data on the same individuals repeatedly, causing a nested structure in the data. Moreover, the data collection is never 'finished' as new data keep streaming in. It is well known that predictions that use the data of the individual whose individual-level effect is predicted in combination with the data of all the other individuals, are better in terms of squared error than those that just use the individual mean. However, when data are both nested and streaming, and the outcome variable is binary, computing these individual-level predictions can be computationally challenging. Five computationally-efficient estimation methods which do not revise "old" data but do account for the nested data structure are developed and evaluated. The methods are based on existing shrinkage factors. A shrinkage factor is used to predict an individual-level effect (i.e., the probability to score a 1), by weighing the individual mean and the mean over all data points. The performance of the existing and newly developed shrinkage factors are compared in a simulation study. While the existing methods differ in their prediction accuracy, the differences in accuracy between the novel shrinkage factors and the existing methods are extremely small. The novel methods are however computationally much more appealing. (C) 2019 Elsevier B.V. All rights reserved.

Chemotherapy is a routine treatment approach for early-stage cancers, but the effectiveness of such treatments is often limited by drug resistance, toxicity, and tumor heterogeneity. Combination chemotherapy, in which two or more drugs are applied simultaneously, offers one promising approach to address these concerns, since two single-target drugs may synergize with one another through interconnected biological processes. However, the identification of effective dual therapies has been particularly challenging; because the search space is large, combination success rates are low. Here, we present our method for DREAM AstraZeneca-Sanger Drug Combination Prediction Challenge to predict synergistic drug combinations. Our approach involves using biologically relevant drug and cell line features with machine learning. Our machine learning model obtained the primary metric = 0.36 and the tie-breaker metric = 0.37 in the extension round of the challenge which was ranked in top 15 out of 76 submissions. Our approach also achieves a mean primary metric of 0.39 with ten repetitions of 10-fold cross-validation. Further, we analyzed our model’s predictions to better understand the molecular processes underlying synergy and discovered that key regulators of tumorigenesis such as TNFA and BRAF are often targets in synergistic interactions, while MYC is often duplicated. Through further analysis of our predictions, we were also ble to gain insight into mechanisms and potential biomarkers of synergistic drug pairs.

PubMed: 31209238, The effectiveness of most cancer targeted therapies is short-lived. Tumors often develop resistance that might be overcome with drug combinations. However, the number of possible combinations is vast, necessitating data-driven approaches to find optimal patient-specific treatments. Here we report AstraZeneca’s large drug combination dataset, consisting of 11,576 experiments from 910 combinations across 85 molecularly characterized cancer cell lines, and results of a DREAM Challenge to evaluate computational strategies for predicting synergistic drug pairs and biomarkers. 160 teams participated to provide a comprehensive methodological development and benchmarking. Winning methods incorporate prior knowledge of drug-target interactions. Synergy is predicted with an accuracy matching biological replicates for >60% of combinations. However, 20% of drug combinations are poorly predicted by all methods. Genomic rationale for synergy predictions are identified, including ADAM17 inhibitor antagonism when combined with PIK3CB/D inhibition contrasting to synergy when combined with other PI3K-pathway inhibitors in PIK3CA mutant cells. © 2019, The Author(s)., National Institute for Health Research, NIHR Wellcome Trust, WT: 102696, 206194 Magyar Tudományos Akadémia, MTA Bayer 668858 PrECISE AstraZeneca, We thank the Genomics of Drug Sensitivity in Cancer and COSMIC teams at the Wellcome Trust Sanger Institute for help with the preparation of the molecular data, Denes Turei for help with Omnipath, and Katjusa Koler for help with matching drug names across combination screens. We thank AstraZeneca for funding and provision of data to the DREAM Consortium to run the challenge, and funding from the European Union Horizon 2020 research (under grant agreement No 668858 PrECISE to J.S.R.), the Joint Research Center for Computational Biomedicine (which is partially funded by Bayer AG) to J.S.R., National Institute for Health Research (NIHR) Sheffield Biomedical Research Center, Premium Postdoctoral Fellowship Program of the Hungarian Academy of Sciences. M.G lab is supported by Wellcome Trust (102696 and 206194)., Competing interests: K.C.B., Z.G., G.Y.D., E.K.Y.T., S.F., and J.R.D. are AstraZeneca employees. K.C.B., Z.G., E.K.Y.T., S.F., and J.R.D. are AstraZeneca shareholders. Y.G. receives personal compensation from Eli Lilly and Company, is a shareholder of Cleerly, Inc., and Ann Arbor Algorithms, Inc. M.G. receives research funding from AstraZeneca and has performed consultancy for Sanofi. The remaining authors declare no competing interests.

As part of the Biomedical Data Translator program, the National Center for Advancing Translational Sciences (NCATS) has assembled 11 teams comprising nearly 200 team members drawn from 28 institutions. Roughly 1 year into the feasibility phase, we describe the program's conception, the rapid coalescence of teams and team members, the novel mechanisms of interaction and communication, and the emergent collaborative culture and community that we believe are driving the early success of the program.