Search results

© 2018 IEEE. Personal use of this material is permitted. Permissíon from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertisíng or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. ; [EN] In the last years, Deep Bidirectional Recurrent Neural Networks (DBRNN) and DBRNN with Long Short-Term Memory cells (DBLSTM) have outperformed the most accurate classifiers for confidence estimation in automatic speech recognition. At the same time, we have recently shown that speaker adaptation of confidence measures using DBLSTM yields significant improvements over non-adapted confidence measures. In accordance with these two recent contributions to the state of the art in confidence estimation, this paper presents a comprehensive study of speaker-adapted confidence measures using DBRNN and DBLSTM models. Firstly, we present new empirical evidences of the superiority of RNN-based confidence classifiers evaluated over a large speech corpus consisting of the English LibriSpeech and the Spanish poliMedia tasks. Secondly, we show new results on speaker-adapted confidence measures considering a multi-task framework in which RNN-based confidence classifiers trained with LibriSpeech are adapted to speakers of the TED-LIUM corpus. These experiments confirm that speaker-adapted confidence measures outperform their non-adapted counterparts. Lastly, we describe an unsupervised adaptation method of the acoustic DBLSTM model based on confidence measures which results in better automatic speech recognition performance. ; This work was supported in part by the European Union's Horizon 2020 research and innovation programme under Grant 761758 (X5gon), in part by the Seventh Framework Programme (FP7/2007-2013) under Grant 287755 (transLectures), in part by the ICT Policy Support Programme (ICT PSP/2007-2013) as part of the Competitiveness and ...

Summarization enhances comprehension and is considered an effective strategy to promote and enhance learning and deep understanding of texts. However, summarization is seldom implemented by teachers in classrooms because the manual evaluation requires a lot of effort and time. Although the need for automated support is stringent, there are only a few shallow systems available, most of which rely on basic word/n-gram overlaps. In this paper, we introduce a hybrid model that uses state-of-the-art recurrent neural networks and textual complexity indices to score summaries. Our best model achieves over 55% accuracy for a 3-way classification that measures the degree to which the main ideas from the original text are covered by the summary . Our experiments show that the writing style, represented by the textual complexity indices, together with the semantic content grasped within the summary are the best predictors, when combined. To the best of our knowledge, this is the first work of its kind that uses RNNs for scoring and evaluating summaries. ; This study is part of the RAGE project. The RAGE project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 644187. This publication reflects only the author's view. The European Commission is not responsible for any use that may be made of the information it contains.

Twitter has become the major source of data for the research community working on the social computing domain. The microblogging site receives millions of tweets every day on its platform. Earlier studies have shown that during any disaster, the frequency of tweets specific to an event grows exponentially, and these tweets, if monitored, processed, and analyzed, can contain actionable information relating to the event. However, during disasters, the number of tweets can be in the hundreds of thousands thereby necessitating the design of a semi-automated artificial intelligence-based system that can extract actionable information based on which steps can be taken for effective disaster response. This paper proposes a Twitter-based disaster response system that uses recurrent nets for training a classifier on a disaster specific tweets dataset. The proposed system would enable timely dissemination of information to various stakeholders so that timely response and proactive measures can be taken in order to reduce the severe consequences of disasters. Experimental results show that the recurrent nets outperform the traditional machine learning algorithms with regard to accuracy in classifying disaster-specific tweets.

Natural outbreaks of transboundary infectious diseases and pandemics are global threats posing international challenges of medical, veterinary, social, and economic character. These diseases have their specific sources and are driven by a range of factors and mechanisms that ensure their transboundary spread. The main driver of transnational spread of infectious diseases is human activity that violates and distorts ecological and climate balance. This disbalance leads to emergence of new pathogens and to expansion of geographical areas of already known diseases and of the range of their host organisms that increasingly include humans. Understanding these aspects is critical for countering existing and future outbreaks of transboundary infections. There is also a risk that infectious potential of microorganisms may be used by armed actors, including parties to politicalmilitary conflicts and terrorists, for their own purposes. While emergence and spread of transboundary infections give rise to a number of problems that reduce the effectiveness of measures for preventing and eliminating them, adequate knowledge about transboundary infections makes it possible to develop a strategy for the management of such diseases at the international level.

Cover -- Title -- Copyright -- Contents -- Series Editors' Introduction -- Preface -- 1 Overview -- 1.1 Introduction -- 1.2 Describing the Events -- 1.3 Using the Event Indicators ("States") -- 1.4 Prediction of Recurrent Events -- 1.5 Conclusion -- 2 Methods for Describing Oscillations, Fluctuations, and Cycles in Univariate Series -- 2.1 Introduction -- 2.2 Types of Movements in Real and Financial Series -- 2.3 Prescribed Rules for Dating Business Cycles -- 2.4 Prescribed Rules for Dating Other Types of Real Cycles -- 2.5 Prescribed Rules for Dating Financial Cycles -- 2.6 Relations between Cycles and Oscillations -- 2.7 The Nature of St and Its Modeling -- 2.8 Conclusion -- 3 Constructing Reference Cycles with Multivariate Information -- 3.1 Introduction -- 3.2 Determining the Reference Cycle via Phases -- 3.3 Combining Specific Cycle Turning Points -- 3.4 Finding Turning Points by Series Aggregation -- 3.5 Conclusion -- 4 Model-Based Rules for Describing Recurrent Events -- 4.1 Introduction -- 4.2 Dating Cycles with Univariate Series -- 4.3 Model-Based Rules for Dating Events with Multivariate Series -- 4.4 Conclusion -- 5 Measuring Recurrent Event Features in Univariate Data -- 5.1 Introduction -- 5.2 The Fraction of Time Spent in Expansions -- 5.3 Representing the Features of Phases -- 5.4 Amplitudes and Durations of Phases -- 5.5 The Shapes of Phases -- 5.6 The Diversity of Phases -- 5.7 Plucking Effects and Recovery Times -- 5.8 Duration Dependence in Phases -- 5.9 Conclusion -- 6 Measuring Synchronization of Recurrent Events in Multivariate Data -- 6.1 Introduction -- 6.2 Moment-Based Measures -- 6.3 Other Approaches to Measuring Synchronization -- 6.4 Synchronization and Model-Based Rules -- 6.5 Application to Synchronization of Industrial Production Cycles -- 6.6 Multivariate Synchronization -- 6.7 Comovement of Cycles.

In 2017, the exacerbation of an ongoing countrywide cholera outbreak in the Democratic Republic of the Congo resulted in >53,000 reported cases and 1,145 deaths. To guide control measures, we analyzed the characteristics of cholera epidemiology in DRC on the basis of surveillance and cholera treatment center data for 2008–2017. The 2017 nationwide outbreak resulted from 3 distinct mechanisms: considerable increases in the number of cases in cholera-endemic areas, so-called hot spots, around the Great Lakes in eastern DRC; recurrent outbreaks progressing downstream along the Congo River; and spread along Congo River branches to areas that had been cholera-free for more than a decade. Case-fatality rates were higher in nonendemic areas and in the early phases of the outbreaks, possibly reflecting low levels of immunity and less appropriate prevention and treatment. Targeted use of oral cholera vaccine, soon after initial cases are diagnosed, could contribute to lower case-fatality rates.

In 2017, the exacerbation of an ongoing countrywide cholera outbreak in the Democratic Republic of the Congo resulted in >53,000 reported cases and 1,145 deaths. To guide control measures, we analyzed the characteristics of cholera epidemiology in DRC on the basis of surveillance and cholera treatment center data for 2008–2017. The 2017 nationwide outbreak resulted from 3 distinct mechanisms: considerable increases in the number of cases in cholera-endemic areas, so-called hot spots, around the Great Lakes in eastern DRC; recurrent outbreaks progressing downstream along the Congo River; and spread along Congo River branches to areas that had been cholera-free for more than a decade. Case-fatality rates were higher in nonendemic areas and in the early phases of the outbreaks, possibly reflecting low levels of immunity and less appropriate prevention and treatment. Targeted use of oral cholera vaccine, soon after initial cases are diagnosed, could contribute to lower case-fatality rates.

Given its size, public procurement matters for economic development. Transparency, competition, accountability, efficiency, and innovation are most commonly noted as guiding principles for achieving best value for money in public contracts. Yet, large-scale, frequently updated, and comparable data on public procurement processes are scarce. This paper presents the methodology and findings of a new global indicator that benchmarks public procurement regulations and practices across 191 economies. The indicator proposes three dimensions to measure the effective implementation of public procurement systems in practice, as applied to a standardized recurrent infrastructure (roads) contract. The three dimensions include the steps and associated time required to complete the procurement process, and the availability and sophistication of e-procurement platforms. A final, fourth component benchmarks the regulatory framework applicable to such contracts. Economies that score higher in the indicator are those with more effective governments, higher quality of roads, and smaller perceptions of corruption. Looking more closely at the scores along the four dimensions reveals that countries differ to a lesser extent in terms of regulatory practices, compared with the use of new technologies such as e-procurement, where considerable gaps between economies exist.

On land transportation vehicles play major role in governmental, nongovernmental organization and individuals for moving goods and peoples. Which used to increase, economy development on the individuals and nations, quality of education, healthcare and other benefit's. But due to vehicles accident countries loss big amount of capital, manpower, and additionally influence negative impact on the healthcare. Vehicle accident injures all age groups. This research is done to provide way to solve this existing problem through sensor fusion. Information fusion provides an important role of any system electrical, electromechanical, biomedical, and any other filed of science and technology. In this research, the main parameters are heartbeat, body temperature, speed and load of the vehicle. The mathematical models of heartbeat are selected by comparing different models from past research [1], [2], [3], [4] and [5]. And the model of body temperature is also selected from past research [6], [7] and [8]. The speed is investigated based on Newton second law of motion on upward, down ward, and on the strait line. Since most of Loads of the vehicle are big, four active metallic strain gauges are used to measure the load of the vehicle. Body temperature of the driver, speed and load of the vehicle are filtered by extended kalman filter. Due to the hard nonlinearity property of heartbeat, it did not filter by extended kalman filter. Then dynamic and linear artificial neural network are designed to track the biological status of the driver and vehicle dynamics respectively. Finally, hierarchical structure, of Bayesian sensor fusion technique is used. These are checked by three Coues that are low, medium and high risk of accident depending on the average traffic accident rule and regulation

Objective:To evaluate the effects of compassion–mindfulness therapy (C-MT), an adapted version of mindfulness-based cognitive therapy that integrates compassion training.Method:Individuals aged 17–69 with recurrent depressive and anxiety symptoms were recruited from a community mental health service unit. Half of the participants were randomized to an 8-week C-MT program ( n = 41) and the other half to a wait-list control condition ( n = 41).Results:Intent-to-treat analyses showed significant improvements in all measures in the treatment group. The effect sizes for depression and anxiety were 1.11 and 1.10, respectively, and those for physical distress, daily functioning, positive affect, and negative affect ranged from 0.71 to 1.04. All improvements were sustained at the 3-month follow-up.Conclusions:The results provide preliminary support for C-MT as a viable treatment option for individuals with recurrent depression and anxiety symptoms. Time-limited treatments such as C-MT should be promoted in social work practice.

Accurate air quality forecasts can provide data-driven supports for governmental departments to control air pollution and further protect the health of residents. However, existing air quality forecasting models mainly focus on site-specific time series forecasts at a local level, and rarely consider the spatiotemporal relationships among regional monitoring stations. As a novelty, we construct a diffusion convolutional recurrent neural network (DCRNN) model that fully considers the influence of geographic distance and dominant wind direction on the regional variations in air quality through different combinations of directed and undirected graphs. The hourly fine particulate matter (PM2.5) and ozone data from 123 air quality monitoring stations in the Yangtze River Delta, China are used to evaluate the performance of the DCRNN model in the regional prediction of PM2.5 and ozone concentrations. Results show that the proposed DCRNN model outperforms the baseline models in prediction accuracy. Compared with the undirected graph model, the directed graph model considering the effects of wind direction performs better in 24 h predictions of pollutant concentrations. In addition, more accurate forecasts of both PM2.5 and ozone are found at a regional level where monitoring stations are distributed densely rather than sparsely. Therefore, the proposed model can assist environmental researchers to further improve the technologies of air quality forecasts and could also serve as tools for environmental policymakers to implement pollution control measures.

Measures of democracy are in high demand. Scientific and public audiences use them to describe political realities and to substantiate causal claims about those realities. This introduction to the thematic issue reviews the history of democracy measurement since the 1950s. It identifies four development phases of the field, which are characterized by three recurrent topics of debate: (1) what is democracy, (2) what is a good measure of democracy, and (3) do our measurements of democracy register real-world developments? As the answers to those questions have been changing over time, the field of democracy measurement has adapted and reached higher levels of theoretical and methodological sophistication. In effect, the challenges facing contemporary social scientists are not only limited to the challenge of constructing a sound index of democracy. Today, they also need a profound understanding of the differences between various measures of democracy and their implications for empirical applications. The introduction outlines how the contributions to this thematic issue help scholars cope with the recurrent issues of conceptualization, measurement, and application, and concludes by identifying avenues for future research.

Measures of democracy are in high demand. Scientific and public audiences use them to describe political realities and to substantiate causal claims about those realities. This introduction to the thematic issue reviews the history of democracy measurement since the 1950s. It identifies four development phases of the field, which are characterized by three recurrent topics of debate: (1) what is democracy, (2) what is a good measure of democracy, and (3) do our measurements of democracy register real-world developments? As the answers to those questions have been changing over time, the field of democracy measurement has adapted and reached higher levels of theoretical and methodological sophistication. In effect, the challenges facing contemporary social scientists are not only limited to the challenge of constructing a sound index of democracy. Today, they also need a profound understanding of the differences between various measures of democracy and their implications for empirical applications. The introduction outlines how the contributions to this thematic issue help scholars cope with the recurrent issues of conceptualization, measurement, and application, and concludes by identifying avenues for future research.