Search

In the field of Data Mining for large scale datasets, also known as Big Data Mining, the availability of effective and efficient classifiers is a prime concern. Accurate classification results can be obtained with sophisticated models, e.g. using ensembling approaches and exploiting concepts of fuzzy set theory, but with an high computational cost. The quest for efficiency leads to the adoption of distributed versions of classification algorithms, and in this effort the support of proper cluster computing frameworks can be fundamental. In this paper it is proposed DFRF, a novel distributed fuzzy random forest induction algorithm, based on a fuzzy discretizer for continuous attributes. The described approach, although shaped on the MapReduce programming model, takes advantage of the implicit distribution of the computation provided by the Apache Spark framework. An extensive experimental characterization of the algorithm over Big Datasets, along with a comparison with other state-of-the-art fuzzy classification algorithms, shows that DFRF provides very competitive results; moreover, a scalability study carried out on a small computer cluster shows that the approach is well behaved with respect to an increment in the number of available computing units.

The ability to properly compare RNA secondary conformations is of prominent importance in many RNA structural investigations, and particularly in the study of non-coding RNAs (ncRNAs). Several different metrics have been proposed to catch topological dissimilarities between RNA secondary structures, but so far specific metric features have not been assessed yet against extensive datasets. The choice of proper metrics is also crucial for the characterization of structural ensembles, so to be able to proficiently analyze them. The current availability of large ncRNA databases has made it possible to perform an extensive comparison of different metrics. Correlation analysis has uncovered the relative descriptive power of such metrics, providing indications on their possible practical use in different contexts.

High performance embedded computing has recently become more and more present in devices used in everyday life. A wide variety of applications, from consumer electronics to biomedical systems, require building up powerful yet cheap embedded devices. Embedded software has tumed out to be more and more complex, posing new challenging issues: the adoption of further flexible programming paradigms/architectures is becoming almost mandatory. Nonetheless, embedded systems development nowadays must rely on a tight coupling of hardware and software components. In this scenario, efficient solutions can be proposed at different levels of abstraction, making use of an assortment of tools and methodologies: researchers and practitioners have a chance to propose new ideas and to compare experimentations. The focus of this conference track is on the application of both novel and well-known techniques to the embedded systems development. Particular attention is paid to solutions that require expertise in different fields (e.g. computer architecture, OS, compilers, security, software engineering, simulation). The track benefits also from direct experiences in the employment of embedded devices in unconventional application areas, so to show up new challenges in the system design/development process. In this setting, researchers, and practitioners from academia and industry get a chance to keep in touch with problems, open issues and future directions in the field of development of dedicated applications for embedded systems.

Document Management Systems (DMSs) are a key component in modern enterprises. For successful document search and retrieval, an adequate metadata set should be defined in order to describe documents with sufficient detail. However, often a single metadata set is not sufficient throughout the whole DMS, as different document types require different attributes to be properly characterized. In this paper, we introduce ontologies as a modeling technology for structured metadata definition within DMSs. Focusing on the ebXML registry standard, we show an approach to enhance DMSs for semantic content management and then we propose a method to exploit this new capability for automated document characterization.

Embedded systems can be regarded today as one of the most lively research and industrial targets. In this field, the ever-increasing demand for computing power and any sort of system resources continuously challenges state-of-the-art design methodologies and development techniques. The largest part of the most complex issues to be addressed during design and development phases can be related to the need to compromise among tight constrains on performance, memory size, code size, power consumption, appliance weight and dimension, possible real-time behaviors, maintainability, scalability, security, time-to-market and, last but not indubitably least, cost. The designer is continuously challenged by requirements modifications that must be timely addressed, operating at different levels of abstraction, e.g. from the proper adjustment of CPU internals to the scheduler structuring in RTOSs, from the optimized exploitation of I/O devices to the employment of specialized compilation techniques. Suitable solutions to emerging problems in this application field can be found via a joint effort by academia and industry.

Many distributed applications make use of distributed object technology. In this kind of systems, modules providing services are implemented as objects spread over a network. Distributed objects are usually accessed through communication frameworks based on specific middleware solutions, such as CORBA, DCOM, and RMI. Applications of this kind might be built up (or extended) integrating different modules, possibly already coded and available on the market. Each required and available module might use a specific communication framework, hampering its prompt integration into a system exploiting a different framework. A convenient way to tackle this problem is the insertion of a gateway module, passing service requests between two different middleware solutions. This approach allows a quick integration of service modules, but it could lead to performance problems, due to the introduced communication overhead. In this paper, we report our experience in developing a simple CORBA/RMI gateway module, and we discuss how it affects the application performance. Measures of the communication overhead show that the employment of the gateway is a viable solution in many real-world applications, and gives hints for efficiently placing modules on the available hosts.

SaaS application instances typically serve multiple tenants to improve cost-efficiency. This results in the need for adaptive performance isolation between tenants in order to guarantee custom service level objectives (SLOs) about request latency or throughput. Current solutions, which are based on request scheduling algorithms, suffer from SLO instability under globally varying workloads. This means that the configuration for an SLO has to be recalibrated when total workload patterns change such as an increase or decrease in the number of subscribed tenants, or the application becomes co-located with other types of resource-intensive applications. Lately container technology such as Docker and container orchestration frameworks like Kubernetes have been used to increase cost-efficiency, multi-tenancy and elasticity. This paper investigates if the problem of adaptive performance isolation can be mapped to resource management concepts of Kubernetes through a series of experiments. These experiments show that Kubernetes provides good support for QoS differentiation and adaptive resource allocation by grouping tenants according to their SLO class (e.g gold vs bronze) in different containers. Moreover, SLO instability does not occur when co-locating these containers with other container-based applications provided that a few interferences between CPU-, memory- and disk-io intensive applications are taken into account. However SLO instability does occur when the number of subscribed tenants changes. This latter problem is not caused by the replication and auto-scaling concepts of Kubernetes, but by a non-linear resource scaling phenomenon that is inherent when the goal is to meet multiple custom SLOs in a cost-optimal way.

Threat modeling is widely adopted and increasingly recognized as an essential step in the secure software development life cycle (SDLC). Focused on privacy-specific threat categories, LINDDUN is a threat modeling framework that allows the identification of privacy-related design flaws at the stage of the initial architecture concept. LINDDUN advocates making explicit any assumptions during the identification and prioritization of privacy threats. These assumptions are in practice documented informally in a free-form, textual format, and the impact, nature and purpose of these assumptions within the context of LINDDUN is not well understood. We present a descriptive study of assumptions made during the application of LINDDUN. This empirical study involves in total 122 threat models created for an IoT-based home automation system and a total of 845 studied assumptions. This study focuses on (i) clarifying the role of assumption-making in the threat modeling process, and (ii) categorizing the types of information provided in these assumptions, and (iii) their relation to the LINDDUN threat categories or more broadly, any privacy-specific concepts. Our results indicate that in practice, (i) assumptions are used to motivate exclusion of potential threats, yet the rationale behind such decisions is often not documented, (ii) many assumption are about the system under analysis, and (iii) a majority is also relevant outside of the specific LINDDUN or privacy scope.

Our main use case features multiple companies that iteratively optimize on the architectural properties of aircraft components in a decentralized manner. In each optimization step of the so-called multi-disciplinary optimization (MDO) process, sensitive data is exchanged between organizations, and we require auditability and traceability of actions taken to assure compliance with signed legal agreements. In this paper, we present a distributed protocol that coordinates authenticated and auditable exchanges of files, leveraging a smart contract. The entire life cycle of a file exchange, including file registration, access request and key distribution, is recorded and traceable via the smart contract. Moreover, when one party raises a dispute, the smart contract can be used to identify the dishonest party without compromising the file's confidentiality. The proposed protocol provides a simple, novel, yet efficient approach to exchange files with support for data access auditability between companies involved in a private consortium with no incentive to share files outside of the protocol. We implemented the protocol in Solidity, deployed it on a private Ethereum blockchain, and validated it within the use case of a decentralized workflow.