Search

The usage of non-traditional data in credit scoring, from microfinance institutions, is very useful when trying to address the problem, very common in emerging markets, of the lack of a verifiable customers’ credit history. In this context, this paper relies on data acquired from smartphones in a loan classification problem. We conduct a set of experiments concerning feature selection, strategies to deal with imbalanced datasets and algorithm choice, to define a baseline model. This model is, then, compared to others adding network features to the original ones. For that comparison, we generate a network that links a given user to its phone book contacts which are users of a given mobile application, taking into account the ethics and privacy concerns involved, and use some feature extraction techniques, such as the introduction of centrality measures and the definition of node embeddings, in order to capture certain aspects of the network’s topology. Several node embedding algorithms are tested, but only Node2Vec proves to be significantly better than the baseline model, applying Friedman’s post hoc tests. This node embedding algorithm outperforms all the other, representing a relative improvement, in comparison with the baseline model, of 5.74% on the mean accuracy, 7.13% on the area under the Receiver Operating Characteristic curve and 30.83% on the Kolmogorov–Smirnov statistic scores. This method, therefore, proves to be very promising when trying to discriminate between “good” and “bad” customers, in credit scoring classification problems.

Although graph-databases have been assuming an increasing relevance in applications that exhibit strong dependability requirements, including tolerance to malicious faults, few works have addressed Byzantine fault tolerance in this particular context, and previous attempts suffer from lack of flexibility and poor performance. This article describes and evaluates Fireplug, a flexible architecture to build robust geo-replicated graph databases. Fireplug can be configured to tolerate from crash to Byzantine faults, both within and across different datacenters. Furthermore, Fireplug is robust to bugs in existing graph database implementations, as it allows to combine multiple graph database instances in a cohesive manner. Thus, Fireplug can support many different deployments, according to the performance/robustness trade-offs imposed by the target application. Our evaluation shows that Fireplug is able implement Byzantine fault tolerance without penalty when compared to the built-in replication mechanism of Neo4j, which only supports crash faults. Additionally, performance optimizations introduced by Fireplug improve the overall performance by up to 900 percent in geo-replicated scenarios.

The largest bottleneck to the development of convolutional neural network (CNN) models in the computational pathology domain is the collection and curation of diverse training datasets. Training CNNs requires large cohorts of image data, and model generalizability is dependent on training data heterogeneity. Including data from multiple centers enhances the generalizability of CNN based models, but this is hindered by the logistical challenges of sharing medical data. In this paper we explore the feasibility of training our recently developed cloud-based segmentation tool (Histo-Cloud) using federated learning. We show that a federated trained model to segment interstitial fibrosis and tubular atrophy (IFTA) using datasets from three institutions is comparable to a model trained by pooling the data on one server when tested on a fourth (holdout) institution’s data. Further, training a model to segment glomeruli for a federated dataset (split by staining) demonstrates similar performance.

Inspection of physical memory allows digital investigators to retrieve evidence otherwise inaccessible when analyzing other storage media. In this paper, we analyze in-memory communication records produced by instant messaging and email applications, both in desktop web-based applications and native applications running in mobile devices. Our results show that, in spite of the heterogeneity of data formats specific to each application, communication records can be represented in a common application-independent format. This format can then be used as a common representation to allow for general analysis of digital artifacts across various applications. Then, we introduce RAMAS, an extensible forensic tool which aims to ease the process of analysing communication records left behind in physical memory by instant-messaging and email clients.

Objective: to build assistive technology, in video format, for deaf and hearing people about sexual health and condom use. Methods: methodological study consisting of the stages of pre-production, production and post-production. In pre-production we built the script, storyboard and scenes, following the principles of Lev Semenovich Vygotsky’s Theory of Learning. In production there was the development of the animation and recording of the scenes by the Libras interpreter. In post-production the final editing of the video took place. Results: the video “Condom: learn to use correctly”,lasting nine minutes and 25 seconds discusses the sexual and reproductive system of man and woman; mode of action, benefits and care with the use of condoms.Given the scarcity of educational materials on the subject, accessibility resources were used to build a video for deaf and hearing people. Conclusion: the construction of the video as an assistive technology directed to the deaf was successfully accomplished. Objetivo: construir tecnologia assistiva, no formato de vídeo, para surdos e ouvintes sobre saúde sexual e o uso do preservativo. Métodos: estudo metodológico constituído das etapas de pré-produção, produção e pós-produção. Na pré-produção construiu-se roteiro, storybord e cenas, seguindo os princípios da Teoria da Aprendizagem de Lev Semenovich Vygotsky. Na produção houve o desenvolvimento da animação e gravação das cenas pelo intérprete de Libras. Na pós-produção ocorreu a edição final do vídeo. Resultados: o vídeo “Preservativo: aprenda a usar corretamente”, com duração de nove minutos e 25 segundos discorre sobre o sistema sexual e reprodutivo do homem e da mulher; modo de ação, benefícios e cuidados com o uso do preservativo. Diante da escassez de materiais educativos sobre o tema, empregaram-se recursos de acessibilidade para construir um vídeo para surdos e ouvintes. Conclusão: a construção do vídeo como tecnologia assistiva dirigida aos surdos foi realizada com êxito.

This paper investigates the problem of integrating two powerful abstractions for concurrent programming, namely futures and transactional memory. Our focus is on specifying the semantics of execution of "transactional futures", i.e., futures that execute as atomic transactions and that are spawned/evaluated by other (plain) transactions or transactional futures. We show that, due to the ability of futures to generate parallel computations with complex dependencies, there exist several plausible (i.e., intuitive) alternatives for defining the isolation and atomicity semantics of transactional futures. The alternative semantics we propose explore different trade-offs between ease of use and efficiency. We have implemented the proposed semantics by introducing a graph-based software transactional memory algorithm, which we integrated with a state of the art JAVA-based Software Transactional Memory (STM). We quantify the performance trade-offs associated with the different semantics using an extensive experimental study encompassing a wide range of diverse workloads.

This paper uses non-traditional data, from a MicroFinance Institution (MFI), in a Credit Scoring loan classification problem and addresses a common problem in emerging markets of the lack of a verifiable customers’ credit history. We perform a set of experiments to define a baseline model and prove the relevance of node embedding features, in credit scoring models, using a real world dataset.

This paper extends the Transactional Memory (TM) paradigm by proposing a new powerful abstraction, the transactional future. Transactional futures, as the name suggests, combine TM with futures, by allowing programmers to exploit intra-transaction parallelism via the abstraction of futures, while delegating to TM the complexity of regulating concurrent access to shared data. The main contribution of this paper is the definition of a set of semantics for transactional futures that explore different trade-offs between simplicity and efficiency.

Made available in DSpace on 2020-12-10T17:33:35Z (GMT). No. of bitstreams: 0 Previous issue date: 2020-05-19 Instituto Chico Mendes de Conservacao da Biodiversidade (ICMBio) Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) Fundacao de Amparo a Pesquisa e Inovacao do Espirito Santo FAPES Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) Background Brazil is one of the most biodiverse countries in the world, with about 37,000 species of land plants. Part of this biodiversity is within protected areas. The development of online databases in the last years greatly improved the available biodiversity data. However, the existing databases do not provide information about the protected areas in which individual plant species occur. The lack of such information is a crucial gap for conservation actions. This study aimed to show how the information captured from online databases, cleaned by a protocol and verified by taxonomists allowed us to obtain a comprehensive list of the vascular plant species from the Parque Nacional do Itatiaia, the first national park founded in Brazil. All existing records in the online database JABOT (15,100 vouchers) were downloaded, resulting in 11,783 vouchers identified at the species level. Overall, we documented 2,316 species belonging to 176 families and 837 genera of vascular plants in the Parque Nacional do Itatiaia. Considering the whole vascular flora, 2,238 species are native and 78 are non-native. New information The Parque Nacional do Itatiaia houses 13% of the angiosperm and 37% of the fern species known from the Brazilian Atlantic Forest. Amongst these species, 82 have been cited as threatened, following IUCN categories (CR, EN or VU), seven are data deficient (DD) and 15 have been classified as a conservation priority, because they are only known from a single specimen collected before 1969. Univ Fed Espirito Santo, Vitoria, ES, Brazil Inst Pesquisas Jardim Bot Rio de Janeiro, Rio De Janeiro, Brazil Univ Estadual Santa Cruz, Ilheus, BA, Brazil Univ Estadual Feira De Santana, Feira De Santana, BA, Brazil Univ Fed Rio de Janeiro, Rio De Janeiro, Brazil Univ Estado Minas Gerais, Belo Horizonte, MG, Brazil Univ Fed Juiz De Fora, Juiz De Fora, Brazil Univ Fed Alfenas, Alfenas, Brazil Inst Bot, Sao Paulo, Brazil Univ Estadual Norte Fluminense, Campos Dos Goytacazes, Brazil Univ Sao Paulo, Sao Paulo, Brazil Univ Fed Parana, Curitiba, Parana, Brazil Embrapa Clima Temperado, Pelotas, RS, Brazil Univ Fed Santa Catarina, Florianopolis, SC, Brazil Univ Fed Pelotas, Pelotas, RS, Brazil Univ Fed Oeste Para, Santarem, Brazil Univ Fed Rural Rio De Janeiro, Seropedica, Brazil Univ Fed Sao Joao Del Rei, Sao Joao Del Rei, Brazil Museu Paraense Emilio Goeldi, Belem, Para, Brazil Univ Estadual Paulista, Sao Paulo, Brazil Flowr Corp, Markham, ON, Canada Univ Estado Rio De Janeiro, Rio De Janeiro, Brazil Univ Estadual Campinas, Campinas, Brazil Univ Fed Rio Grande do Sul, Porto Alegre, RS, Brazil Univ Estadual Paulista, Sao Paulo, Brazil Instituto Chico Mendes de Conservacao da Biodiversidade (ICMBio): 18/2017 CNPq: 18/2017 CNPq: CNPq 421276/2017-7 Fundacao de Amparo a Pesquisa e Inovacao do Espirito Santo: 18/2017 Fundacao de Amparo a Pesquisa e Inovacao do Espirito Santo: FAPES 34/2018 FAPES: 525/2018 FAPERJ: E-26/202.775/2018 FAPERJ: E-26/202.778/2018

Since early 2000, Microfinance Institutions (MFI) have been using credit scoring for their risk assessment. However, one of the main problems of credit scoring in microfinance is the lack of structured financial data. To address this problem, MFI have started using non-traditional data which can be extracted from the digital footprint of their users. The non-traditional data can be used to build algorithms that can identify good borrowers as in traditional banking. This paper proposes an assembled method to evaluate the predictive power of the non-traditional method. By using the Weight of Evidence (WoE), a transformation based on the distribution within the feature, as feature transformation method, and then applying extremely randomized trees for feature selection, we were able to improve the accuracy of the credit scoring model by 20.20% when compared to the credit scoring model built with the traditional implementation of WoE. This paper shows how the assembling of WoE with different feature selection criteria can result in more robust credit scoring models in microfinance.

Background and Objective: Glaucoma, an eye condition that leads to permanent blindness, is typically asymptomatic and therefore difficult to be diagnosed in time. However, if diagnosed in time, Glaucoma can effectively be slowed down by using adequate treatment; hence, an early diagnosis is of utmost importance. Nonetheless, the conventional approaches to diagnose Glaucoma adopt expensive and bulky equipment that requires qualified experts, making it difficult, costly and time-consuming to diagnose large amounts of people. Consequently, new alternatives to diagnose Glaucoma that suppress these issues should be explored. Methods: This work proposes an interpretable computer-aided diagnosis (CAD) pipeline that is capable of diagnosing Glaucoma using fundus images and run offline in mobile devices. Several public datasets of fundus images were merged and used to build Convolutional Neural Networks (CNNs) that perform segmentation and classification tasks. These networks are then used to build a pipeline for Glaucoma assessment that outputs a Glaucoma confidence level and also provides several morphological features and segmentations of relevant structures, resulting in an interpretable Glaucoma diagnosis. To assess the performance of this method in a restricted environment, this pipeline was integrated into a mobile application and time and space complexities were assessed. Results: Considering the test set, the developed pipeline achieved 0.91 and 0.75 of Intersection over Union (IoU) in the optic disc and optic cup segmentation, respectively. With regards to the classification, an accuracy of 0.87 with a sensitivity of 0.85 and an AUC of 0.93 were attained. Moreover, this pipeline runs on an average Android smartphone in under two seconds. Conclusions: The results demonstrate the potential that this method can have in the contribution to an early Glaucoma diagnosis. The proposed approach achieved similar or slightly better metrics than the current CAD systems for Glaucoma assessment while running on more restricted devices. This pipeline can, therefore, be used to construct accurate and affordable CAD systems that could enable large Glaucoma screenings, contributing to an earlier diagnose of this condition.

This paper studies the possibility of using the encrypted video channel of widely used videoconferencing applications, such as Skype, as a carrier for unobservable covert TCP/IP communications. We propose and evaluate different alternatives to encode information in the video stream in order to increase available throughput while preserving the packet-level characteristics of the video stream. We have built a censorship-resistant system, named DeltaShaper, which offers a data-link interface and supports TCP/IP applications that tolerate low throughput / high latency links. Our results show that it is possible to run standard protocols such as FTP, SMTP, or HTTP over Skype video streams.

This paper addresses the problem of offering reliable causal multicast in a setting where nodes are organized in an overlay network and use this network to disseminate information among each other. The use of overlay networks for this purpose is widely used when the number of nodes is large. For instance, many publish-subscribe systems use an overlay of message brokers to support the exchange of information among publishers and subscribers. To the best of our knowledge, previous multicast algorithms for overlay networks either do not enforce causal order or, in order to do so, require nodes to keep metadata (for instance, sequence numbers) for all senders and are, therefore, inherently non-scalable. In this paper we propose a novel localized algorithm to implement reliable causal multicast, where each node is only required to keep metadata regarding nodes in its neighbourhood (with a radius that is a function of the number of faults that need to be tolerated). Experimental results show that our algorithm can achieve significant improvements over non-localized alternatives, and can even outperform localized algorithms that do not offer causal order.

Modern fault-tolerant distributed architectures can be configured to tolerate a wide-range of faults. For instance, Fireplug is a distributed BFT graph database, based on n-version programming, that can be configured to tolerate crash or Byzantine faults, uncorrelated faults in individual machines, correlated faults that affect all replicas running a given software version, or correlated faults that affect an entire datacenter. Interestingly, in such a system, fault handling heavily depends on the type of faults the system is configured to tolerate. To hardwire all possible behaviours in the fault-handling code is inflexible and may even be impractical. In this paper, we explore a different alternative that consists in specifying not only the system configuration, but also the fault-handling behaviour, and how the system adapts to changes in the workload, in a policy language, that is processed externally to the managed system. We show that, using this approach, a single simplified codebase of the managed system can be used effectively to address a wide range of dependability constraints.

Many adaptive systems react to variations in their environment by changing their con guration. Often, they make the adaptation decisions based on some knowledge about how the reconfiguration actions impact the key performance indicators. However, the outcome of these actions is typically affected by uncertainty. Adaptation actions have non-deterministic impacts, potentially leading to multiple outcomes. When this uncertainty is not captured explicitly in the models that guide adaptation, decisions may turn out ineffective or even harmful to the system. Also critical is the need for these models to be interpretable to the human operators that are accountable for the system. However, accurate impact models for actions that result in non-deterministic outcomes are very difficult to obtain and existing techniques that support the automatic generation of these models, mainly based on machine learning, are limited in the way they learn non-determinism. In this paper, we propose a method to learn human-readable models that capture non-deterministic impacts explicitly. Additionally, we discuss how to exploit expert's knowledge to bootstrap the adaptation process as well as how to use the learned impacts to revise models defined offline. We motivate our work on the adaptation of applications in the cloud, typically affected by hardware heterogeneity and resource contention. To validate our approach we use a prototype based on the RUBiS auction application.

The problem of distributed consensus in the presence of Byzantine faults has received particular attention in recent decades. Today a variety of solution to this problem exist, each optimized for particular execution conditions. Given that, in most cases, real systems operate under dynamic conditions, it is important to develop mechanisms that allow the algorithms to be adapted or changed at runtime to optimize the system to the current conditions. The problem of dynamic adaptation of consensus algorithms is not new, but the literature is scarce for the Byzantine case and there is no comprehensive comparison of existing solutions. This work has two complementary objectives. First, it studies how the different dynamic adaptation techniques proposed for the crash failure model can be applied in the presence of Byzantine faults. Second, it presents a comparative study of the performance of these switching algorithms in practice. For that purpose, we have implemented the switching algorithms in a common software framework, based on the open source BFT-SMaRt library. Using this framework we have performed an extensive evaluation that offers useful insights on the practical effects of different mechanisms used to support the run-time switching among Byzantine protocols.

The paper describes and evaluates Fireplug, a flexible architecture to build robust geo-replicated graph databases. Fireplug can be configured to tolerate from crash to Byzantine faults, both within and across different datacenters. Furthermore, Fireplug is robust to bugs in existing graph database implementations, as it allows to combine multiple graph databases instances in a cohesive manner. Thus, Fireplug can support many different deployments, according to the performance/robustness tradeoffs imposed by the target application. Our evaluation shows that Fireplug can implement Byzantine fault tolerance in geo-replicated scenarios and still outperform the built-in replication mechanism of Neo4j, which only supports crash faults.

Replication plays an essential role for in-memory distributed transactional platforms, such as NoSQL data grids, given that it represents the primary mean to ensure data durability. Unfortunately, no single replication technique can ensure optimal performance across a wide range of workloads and system configurations. This paper tackles this problem by presenting MORPHR, a framework that allows to automatically adapt the replication protocol of in-memory transactional platforms according to the current operational conditions. MORPHR presents two key innovative aspects. On one hand, it allows to plug in, in a modular fashion, specialized algorithms to regulate the switching between arbitrary replication protocols. On the other hand, MORPHR relies on state of the art machine learning techniques to autonomously determine the optimal replication in face of varying workloads. We integrated MORPHR in a popular open-source in-memory NoSQL data grid, and evaluated it by means of an extensive experimental study. The results highlight that MORPHR is accurate in identifying the optimal replication strategy in presence of complex, realistic workloads, and does so with minimal overhead.

We take a game theoretical approach to determine necessary and sufficient conditions under which we can persuade rational agents to exchange messages in pairwise exchanges over links of a dynamic network, by holding them accountable for deviations with punishments. We make three contributions: (1) we provide a new game theoretical model of repeated interactions in dynamic networks, where agents have incomplete information of the topology, (2) we define a new solution concept for this model, and (3) we identify necessary and sufficient conditions for enforcing accountability, i.e., for persuading agents to exchange messages in the aforementioned model. Our results are of technical interest but also of practical relevance. We show that we cannot enforce accountability if the dynamic network does not allow for \emph{timely punishments}. In practice, this means for instance that we cannot enforce accountability in some networks formed in file-sharing applications such as Bittorrent\,\cite{Cohen:03}. We also show that for applications such as secret exchange, where the benefits of the exchanges significantly surpass the communication costs, timely punishments are enough to enforce accountability. However, we cannot in general enforce accountability if agents do not possess enough information about the network topology. Nevertheless, we can enforce accountability in a wide variety of networks that satisfy 1-connectivity\,\cite{Kuhn:10} with minimal knowledge about the network topology, including overlays for gossip dissemination such as \cite{Li:06,Li:08}., Comment: 32 pages, 3 figures, 7 main theorems, 1 algorithm, to be submitted to a conference; improved presentation; added generalisation to non-bounded protocols

Hyperspace hashing is a recent multi-dimensional indexing technique for distributed key-value stores that aims at supporting efficient queries using multiple objects' attributes. However, the advantage of supporting complex queries comes at the cost of a complex configuration. In this paper we address the problem of automating the configuration of this innovative distributed indexing mechanism. We first show that a misconfiguration may significantly affect the performance of the system. We then derive a performance model that provides key insights on the behaviour of hyperspace hashing. Based on this model, we derive a technique to automatically and dynamically select the best configuration.

Communication through the Internet raises privacy and confidentiality concerns. Protocols such as HTTPS may be used to protect the communication, but occasionally vulnerabilities that may allow snooping on packet content are discovered. To address this issue, we present MACHETE, an application-layer multi-path communication mechanism that provides additional confidentiality by splitting data streams in different physical paths. MACHETE has to handle two challenges: sending packets over different paths when Internet's routing imposes a single path between pairs of network interfaces; splitting streams of data sent over TCP connections. MACHETE is the first to exploit MultiPath TCP (MPTCP) for security purposes. It leverages overlay networks and multihoming to handle the first challenge and MPTCP to handle the second. MACHETE establishes an overlay network and scatters the data over the available paths, thus reducing the effectiveness of snooping attacks. Mechanisms are provided to select paths based on path diversity.

The authentication of users in legacy web sites via mobile devices is still a challenging problem. Users are required to provide passwords, introducing several vulnerabilities: since strong passwords are hard to memorize, users often use weak passwords that are easy to break, and passwords can be intercepted by malware and stolen. In this paper we propose a novel architecture, named BioALeg, to support secure biometric authentication on legacy websites. Our approach leverages the potential of a Secured Personal Device (SPD), a hardware add-on for mobile phones that is being developed in the context of the PCAS European project. The device offers biometric authentication and secure storage services. BioALeg uses this infrastructure, and a companion web site plugin, to support biometric authentication in legacy web sites that currently use username/password authentication. In order to perform authentication, the smartphone requests a One Time Password (OTP) to the service provider when the user tries to access the service using the SPD. Due to the architecture and implementation of the SPD, the OTP transfer only occurs after the owner of the phone and SPD is correctly authenticated using biometrics. The PCAS infrastructure guarantees that, after the biometric authentication, the user identity is valid and accepted by all components. BioALeg has been implemented as an Android service and integrated with legacy web sites.

This work investigates how to combine two powerful abstractions to manage concurrent programming: Transactional Memory (TM) and futures. The former hides from programmers the complexity of synchronizing concurrent access to shared data, via the familiar abstraction of atomic transactions. The latter serves to schedule and synchronize the parallel execution of computations whose results are not immediately required. While TM and futures are two widely investigated topics, the problem of how to exploit these two abstractions in synergy is still largely unexplored in the literature. This paper fills this gap by introducing Java Transactional Futures (JTF), a Java-based TM implementation that allows programmers to use futures to coordinate the execution of parallel tasks, while leveraging transactions to synchronize accesses to shared data. JTF provides a simple and intuitive semantic regarding the admissible serialization orders of the futures spawned by transactions, by ensuring that the results produced by a future are always consistent with those that one would obtain by executing the future sequentially. Our experimental results show that the use of futures in a TM allows not only to unlock parallelism within transactions, but also to reduce the cost of conflicts among top-level transactions in high contention workloads.

Edge-computing is one of the most promising techniques to leverage the excess capacity that exists at users' premises. Unfortunately, edge-computing may be vulnerable to free-riding, i.e., to nodes that attempt to benefit from the system without providing any service in return. Traditional approaches model free-riders as rational nodes that strive to maximize a utility and apply Game Theory concepts to devise mechanisms that deny any utility gain to nodes that deviate from the protocol. These mechanisms impose significant overheads. This paper proposes a new approach that avoids these overheads, which applies concepts of evolutionary game theory. We propose to devise lightweight mechanisms targeted for the optimistic setting where nodes adopt one of a small number of behaviours. If a small fraction of nodes follows alternative behaviours, then the lightweight mechanism should limit the utility gain of these nodesto control the increase in the number of these nodes. This allows altruistic nodes to detect their presence in time to switch to more robust mechanisms before the system reaches a state where the lightweight mechanisms can no longer cope with the existing behaviours. We apply this approach in the context of edge-assisted streaming and propose the use of carefully crafted symmetric overlays to support message dissemination in an environment with free-riders and white-washers. The topology maintenance procedures of our overlay encourage nodes to maintain stable symmetric links. Leveraging the topological properties of the resulting symmetric overlay, direct reciprocity mechanisms deny any utility gain to free-riders and white-washers, while limiting the utility gain of small fractions of nodes that adopt more sophisticated behaviours.

One of the main challenges to harness the potential of Cloud computing is the design of programming models that simplify the development of large-scale parallel applications and that allow ordinary programmers to take full advantage of the computing power and the storage provided by the Cloud, both of which made available, on demand, in a pay-only-forwhat-you-use pricing model. In this paper, we discuss the use of the Transactional Memory programming model in the context of the cloud computing paradigm, which we refer to as Cloud-TM. We identify where existing Distributed Transactional Memory platforms still fail to meet the requirements of the cloud and of its users, and we point several open research problems whose solution we deem as essential to materialize the Cloud-TM vision.

A flexible way of building modular communication stacks relies on the use of protocol composition. This paper describes a protocol composition framework that simplifies the task of deriving the worst-case response time of a protocol composition from the protocol implementation. In order to derive the worst-case response time of a protocol composition, one needs to capture its event-graph: the event-graph consists of the set of all events processed by each component and the relation between those events. The framework, called RT-Appia, takes a pragmatic approach: instead of requiring the use of domain specific code analysis tools, or dedicated compilers, it simply requires protocol programmers to make explicit which events are processed and produced by each layer, and how these events are related. An interesting aspect of the approach is that the same data structures that are used to simplify the task of computing the worst-case response time of the protocol composition are also used to optimize the performance and to debug the resulting implementation.

The paper proposes SmartFetch, a storage strategy that relies on a combination of techniques aimed at efficiently supporting selective jobs that are only concerned with a subset of the entire dataset in systems such as Hadoop and Spark. We combine the use of an appropriate data-layout with data indexing tools to improve the data access speed and significantly shorten total job execution time. An extensive experimental evaluation of SmartFetch shows that, by avoiding reading irrelevant blocks, it can provide significant speedups when compared to the basic Hadoop and Spark implementations. Further, our system also outperforms other implementations that use several variants of the techniques we have embedded in SmartFetch.

This report summarizes the viewpoints and insights gathered in the Dagstuhl Seminar on Atomicity in System Design and Execution, which was attended by 32 people from four different scientific communities: database and transaction processing systems, fault tolerance and dependable systems, formal methods for system design and correctness reasoning, and hardware architecture and programming languages. Each community presents its position in interpreting the notion of atomicity and the existing state of the art, and each community identifies scientific challenges that should be addressed in future work. In addition, the report discusses common themes across communities and strategic research problems that require multiple communities to team up for a viable solution. The general theme of how to specify, implement, compose, and reason about extended and relaxed notions of atomicity is viewed as a key piece in coping with the pressing issue of building and maintaining highly dependable systems that comprise many components with complex interaction patterns.

Over the last years Transactional Memory (TM) gained growing popularity as a simpler, attractive alternative to classic lock-based synchronization schemes. Recently, the TM landscape has been profoundly changed by the integration of Hardware TM (HTM) in Intel commodity processors, raising a number of questions on the future of TM. We seek answers to these questions by conducting the largest study on TM to date, comparing different locking techniques, hardware and software TMs, as well as different combinations of these mechanisms, from the dual perspective of performance and power consumption. Our study sheds a mix of light and shadows on currently available commodity HTM: on one hand, we identify workloads in which HTM clearly outperforms any alternative synchronization mechanism; on the other hand, we show that current HTM implementations suffer of restrictions that narrow the scope in which these can be more effective than state of the art software solutions. Thanks to the results of our study, we identify a number of compelling research problems in the areas of TM design, compilers and self-tuning.

We consider the following natural problem: n failure-prone servers, communicating synchronously through message passing, must assign themselves one-to-one to n distinct items. Existing literature suggests two possible approaches to this problem. First, model it as an instance of tight renaming in synchronous message-passing systems; for deterministic solutions, a tight bound of Θ(log n) communication rounds is known. Second, model the scenario as an instance of randomized load-balancing, for which elegant sub-logarithmic solutions exist. However, careful examination reveals that known load-balancing schemes do not apply to our scenario, because they either do not tolerate faults or do not ensure one-to-one allocation. It is thus natural to ask if sub-logarithmic solutions exist for this apparently simple but intriguing problem. In this paper, we combine the two approaches to provide a new randomized solution for tight renaming, which terminates in O(log log n) communication rounds with high probability, against a strong adaptive adversary. Our solution, called Balls-into-Leaves, combines the deterministic approach with a new randomized scheme to obtain perfectly balanced allocations. The algorithm arranges the items as leaves of a tree, and participants repeatedly perform random choices among the leaves. The algorithm exchanges information in each round to split the participants into progressively smaller groups whose random choices do not conflict. We then extend the algorithm to terminate early in O(log log f) rounds w.h.p., where f is the actual number of failures. These results imply an exponential separation between deterministic and randomized algorithms for the tight renaming problem in message-passing systems., National Science Foundation (U.S.) (Grant CCF-1217921), National Science Foundation (U.S.) (Grant CCF-1301926), United States. Dept. of Energy (Advanced Scientific Computing Research Grant ER26116/DE-SC0008923), Oracle Corporation, Intel Corporation

The advent of multi-core architectures has brought concurrent programming to the forefront of software development. In this context, Transactional Memory (TM) has gained increasing popularity as a simpler, attractive alternative to traditional lock-based synchronization. The recent integration of Hardware TM (HTM) in the last generation of Intel commodity processors turned TM into a mainstream technology, raising a number of questions on its future and that of concurrent programming. To evaluate the potential impact of Intel's HTM, we conducted the largest study on TM to date, comparing different locking techniques, hardware and software TMs, as well as different combinations of these mechanisms, from the dual perspective of performance and power consumption. As a result we perform a workload characterization, to help programmers better exploit the currently available TM facilities, and identify important research directions.

In this paper we present Roller chain, a novel Distributed Hash Table that offers efficient data storage through the combination of gossip-based and structured overlay networks. The unstructured component maintains clusters of fully connected nodes, where each cluster acts as a virtual node in the structured component. This architecture simplifies the management of data replication and balances the load among nodes in the system. We have implemented a prototype of Roller chain that we have used to experimentally validate its performance against other state of the art solutions.

This work uses Game Theory to study the effectiveness of punishments as an incentive for rational nodes to follow an epidemic dissemination protocol. The dissemination process is modeled as an infinite repetition of a stage game. At the end of each stage, a monitoring mechanism informs each player of the actions of other nodes. The effectiveness of a punishing strategy is measured as the range of values of the benefit-to-cost ratio that sustain cooperation. This paper studies both public and private monitoring. Under public monitoring, we show that direct reciprocity is not an effective incentive, whereas full indirect reciprocity provides a nearly optimal effectiveness. Under private monitoring, we identify necessary conditions regarding the topology of the graph in order for punishments to be effective. When punishments are coordinated, full indirect reciprocity is also effective under private monitoring.

The aim of this research was to identify the supervisor's profile of the Portuguese sign language interpreters students that improves and develops their knowledge and skills. We conducted a qualitative study using different methods to collect data, such as interviews and a modified Delphi technique, for which we created an inventory with several competencies based on a literature review. Given the specificity of this area, we felt the need to conceptualize and create a new competence – “gesture performance” defining its specificity. With this paper we pretend to describe the Delphi technique and publicize its results. info:eu-repo/semantics/publishedVersion

This paper addresses the self-management of in-memory distributed data grid platforms. A growing number of applications rely in these platforms to speed up access to large sets of data. However, they are complex to manage due to the diversity of configuration and load profiles. The proposed approach employs an adaptation policy expressed in terms of high-level goals to facilitate the task of the system manager, and address the complexity issues posed by the management of multiple configurations. The approach is validated experimentally using the open-source RedHat´s Infinispan platform.

Multicast is an important building block for many applications in MANETs, including data dissemination, service discovery, publish-subscribe, among others. Therefore, it has been widely studied and many solutions can be found in the literature. However, most existing multicast protocols are tailored to a specific type of mobility pattern and therefore are unable to excel in face of heterogeneous topology stability conditions. This paper proposes and evaluates TSAMP, a Topology Stability-Aware Multicast Protocol for MANETs that exploits locally perceived mobility conditions in order to promote the use of stable routes, when available.

This paper introduces the N-party BAR Transfer service. The service aims at transferring data from a group of replicated producers to a group of replicated consumers, in a setting where the behaviour of the processes is characterized by the Byzantine Altruistic Rational (BAR) model. The problem is motivated, a precise definition of problem is presented, and the challenges on solving the problem are discussed.

Parallel programming (PP) used to be an area once confined to scientific and high-performance computing applications. However, with the proliferation of multicore processors, parallel programming has definitely become a mainstream concern. Transactional Memories (TMs) answer the need to find a better programming model for PP, capable of boosting developers' productivity and allowing ordinary programmers to unleash the power of parallel and distributed architectures avoiding the pitfalls of manual, lock based synchronization.

In-memory NoSQL transactional data grids are emerging as an attractive alternative to conventional relational distributed databases. In these platforms, replication plays a role of paramount importance, as it represents the key mechanism to ensure data durability. In this work we focus on Atomic Broadcast (AB) based certification replication schemes, which have recently emerged as a much more scalable alternative to classical replication protocols based on active replication or atomic commit protocols. We first show that, among the existing AB-based certification protocols, no "one-fits-all" solution exists that achieves optimal performance in presence of heterogeneous workloads. Next, we present PolyCert, a polymorphic certification protocol that allows for the concurrent coexistence of different certification protocols, relying on machine-learning techniques to determine the optimal certification scheme on a per transaction basis. We design and evaluate two alternative oracles, based on parameter-free machine learning techniques that rely both on off-line and on-line training approaches. Our experimental results demonstrate the effectiveness of the proposed approach, highlighting that PolyCert is capable of achieving a performance extremely close to that of an optimal non-adaptive certification protocol in presence of non heterogeneous workloads, and significantly outperform any non-adaptive protocol when used with realistic, complex applications that generate heterogeneous workloads.

Total Order Broadcast (TOB) is a fundamental building block at the core of a number of strongly consistent, fault-tolerant replication schemes. While it is widely known that the performance of existing TOB algorithms varies greatly depending on the workload and deployment scenarios, the problem of how to forecast their performance in realistic settings is, at current date, still largely unexplored. In this paper we address this problem by exploring the possibility of leveraging on machine learning techniques for building, in a fully decentralized fashion, performance models of TOB protocols. Based on an extensive experimental study considering heterogeneous workloads and multiple TOB protocols, we assess the accuracy and efficiency of alternative machine learning methods including neural networks, support vector machines, and decision tree-based regression models. We propose two heuristics for the feature selection phase, that allow to reduce its execution time up to two orders of magnitude incurring in a very limited loss of prediction accuracy.

Quorums are a fundamental building block for solving various fundamental problems such as consensus, distributed dictionaries, distributed storage, among others. In particular, probabilistic quorums have shown to be scalable, efficient, and suitable for dynamic environments [12]. Unfortunately, most existent analytic results for accessing probabilistic quorums are tailored to static networks [8]. However, we believe that the correct functioning of such systems must be assured in a much wider range of scenarios.In this paper, we discuss the random walk based scheme for accessing probabilistic quorums in dynamic networks where an oblivious adversary changes the communication links arbitrarily in each round. We show that O(n3 log n) communication steps are needed to access a probabilistic quorum under these conditions.

In recent years the world has witnessed many catastrophic events where the intervention of first responders was required to manage massive disaster scenarios. To be effective, emergency management teams must be able to coordinate and communicate efficiently. To the best of our knowledge no solution has been proposed to support the communication and coordination of these teams in an integrated fashion. In this paper we propose MUSTUS, an architecture that provides communication mechanisms and coordination primitives, based on the interaction through a tuple space, whose semantics are specifically designed to support the needs of emergency management applications.

Gossip, or epidemic, protocols have emerged as a highly scalable and resilient approach to implement several application level services such as reliable multicast, data aggregation, publish-subscribe, among others. All these protocols organize nodes in an unstructured random overlay network. In many cases, it is interesting to bias the random overlay in order to optimize some efficiency criteria, for instance, to reduce the stretch of the overlay routing. In this paper we propose X-BOT, a new protocol that allows to bias the topology of an unstructured gossip overlay network. X-BOT is completely decentralized and, unlike previous approaches, preserves several key properties of the original (non-biased) overlay (most notably, the node degree and consequently, the overlay connectivity). Experimental results show that X-BOT can generate more efficient overlays than previous approaches., (undefined)

The OSGi Service Platform defines a framework for the deployment of extensible and downloadable Java applications. Many of the application areas for OSGi have significant dependability requirements. This paper presents and evaluates FT-OSGi, a set of extensions to the OSGi Service Platform that allow to replicate OSGi services. FT-OSGi supports replication of OSGi services, including state-transfer among replicas, supports multiple replication strategies, and allows to apply a different replication strategy to each OSGi service.

The transition to multicore architectures has raised the urge to identify novel programming paradigms aimed at simplifying the development of parallel programs.