Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 5
Encontrada(s) 1 página(s)

e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Artículo científico (JournalArticle)

Energy-optimal collaborative file distribution in wired networks

  • Verma, Kshitiz
  • Rizzo, Gianluca
  • Fernández Anta, Antonio
  • Cuevas Rumín, Rubén
  • Azcorra Saloña, Arturo
  • Zaks, Shmuel
  • García Martínez, Alberto
The impact of the ICT sector in worldwide power consumption is an increasing concern, motivating the research community to devote an important effort to define novel energy efficient networking solutions. Despite file distribution is responsible for a major portion of the current Internet traffic, little effort has been dedicated to address the issue of its energy efficiency so far. Most of the previous literature focuses on optimizing the download time of file distribution schemes (e.g. centralized server-based or distributed peer-to-peer solutions) while it is yet unclear how to optimize file distribution schemes from the point of view of energy consumed. In this paper, we present a general modelling framework to analyze the energy consumption of file distribution systems. First, we show that the general problem of minimizing energy consumption in file distribution is NP-hard. Then, for restricted versions of the problem, we establish theoretical bounds to minimal energy consumption. Furthermore, we define a set of optimal algorithms for a variety of system settings, which exploit the service capabilities of hosts in a P2P fashion. We show that our schemes are capable of reducing at least 50 % of the energy consumed by traditional (yet largely used) centralized distribution schemes even when considering effects such as network congestion and heterogeneous access speed across nodes., Supported in part by Ministerio de Economia y Competitividad grant TEC2014- 55713-R, the DRONEXT project (TEC2014-58964-C2-1-R), Regional Government of Madrid (CM) grant Cloud4BigData (S2013/ICE-2894, co- funded by FSE & FEDER), and BRADE Project (P2013/ICE-2958), NSF of China grant 61520106005, and European Commission H2020 grants ReCred and NOTRE.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Artículo científico (JournalArticle)

A taxonomy generation tool for semantic visual analysis of large corpus of documents

  • Carrión Recio, Belén
  • Onorati, Teresa
  • Díaz, Paloma
  • Vasiliki, Triga
Taxonomies are semantic resources that help to categorize and add meaning to data. In a hyperconnected world where information is generated at a rate that exceeds human capacities to process and make sense of it, such semantic resources can help to access relevant information more efficiently by extracting knowledge from large and unstructured data sets. Taxonomies are related to specific domains of knowledge in which they identify relevant topics. However, they have to be validated by experts to guarantee that its terms and relations are meaningful. In this paper, we introduce a semiautomatic taxonomy generation tool for supporting domain experts in building taxonomies that are then used to automatically create semantic visualizations of data. Our proposal combines automatic techniques to extract, sort and categorize terms, and empowers domain experts to take part at any stage of the process by providing a visual edition tool. We tested the tool's usability in two use cases from different domains and languages. Results show that all the functionalities are easy to use and interact with. Lessons learned from this experience will guide the design of a utility evaluation involving domain experts interested in data analysis and knowledge modeling, This work was supported by the project PACE grant funded by the Spanish Ministry of Economy and Competitivity [TIN2016- 77690-R]. Authors also acknowledge travel funding from the European Union’s Horizon 2020 Framework through NOTRE project (H2020-TWINN-2015, GA Number: 692058).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publicación Contribución de congreso (ConferenceOutput)

Inclusive access to emergency services: An action research project focused on hearing impaired citizens

  • Constantinou, Vaso
  • Ioannou, Andri
  • Díaz, Paloma
Poster Papers of ISCRAM-Med 2016 - Third International Conference on Information Systems for Crisis Response and Management in Mediterranean Countries. October 26-28, 2016. Universidad Carlos III de Madrid (Spain), In case of emergency, hearing impaired people are not always able to access emergency services and, hence, they do not have equal access to social support and infrastructure. In this work we undertake the development and evaluation of a system aiming to meet the communication needs of hearing impaired citizens in cases of emergency. The system consists of (i) a mobile application that records and sends the details of an emergency event, and (ii) a central management system that handles these calls from the operation center at the emergency services. The system was completed in four cycles of design, development and evaluation with the involvement of 74 hearing impaired users and three officers from the Cyprus Police (Emergency Response Unit)., Authors acknowledge travel funding from the European Union's Horizon 2020 Framework through NOTRE project (H2020-TWINN-2015, GA Number: 692058).
Dipòsit Digital de la UB
Publicación Artículo científico (JournalArticle)

Geometric Correlations Mitigate the Extreme Vulnerability of Multiplex Networks against Targeted Attacks

  • Kleineberg, Kaj Kolja
  • Buzna, Lubos
  • Papadopoulos, Fragkiskos
  • Boguñá, Marián
  • Serrano Moral, Ma. Ángeles (María Ángeles)
We show that real multiplex networks are unexpectedly robust against targeted attacks on high-degree nodes and that hidden interlayer geometric correlations predict this robustness. Without geometric correlations, multiplexes exhibit an abrupt breakdown of mutual connectivity, even with interlayer degree correlations. With geometric correlations, we instead observe a multistep cascading process leading into a continuous transition, which apparently becomes fully continuous in the thermodynamic limit. Our results are important for the design of efficient protection strategies and of robust interacting networks in many domains.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Geometric Correlations Mitigate the Extreme Vulnerability of Multiplex Networks against Targeted Attacks

  • Kleineberg, Kaj Kolja
  • Buzna, Lubos
  • Papadopoulos, Fragkiskos
  • Boguñá, Marián
  • Serrano Moral, Ma. Ángeles (María Ángeles)
We show that real multiplex networks are unexpectedly robust against targeted attacks on high-degree nodes and that hidden interlayer geometric correlations predict this robustness. Without geometric correlations, multiplexes exhibit an abrupt breakdown of mutual connectivity, even with interlayer degree correlations. With geometric correlations, we instead observe a multistep cascading process leading into a continuous transition, which apparently becomes fully continuous in the thermodynamic limit. Our results are important for the design of efficient protection strategies and of robust interacting networks in many domains.