2014
Contribution to conference
Open Access
LearNext: learning to predict tourists movements
Baraglia R., Muntean C. I., Nardini F. M., Silvestri F.In this paper, we tackle the problem of predicting the "next" geographical position of a tourist given her history (i.e., the prediction is done accordingly to the tourist's current trail) by means of supervised learning techniques, namely Gradient Boosted Regression Trees and Rank- ing SVM. The learning is done on the basis of an object space represented by a 68 dimension feature vector, specifically designed for tourism related data. Furthermore, we propose a thorough comparison of several methods that are considered state-of-the-art in touristic recommender and trail prediction systems as well as a strong popularity baseline. Experiments show that the methods we propose outperform important competitors and baselines thus providing strong evidence of the performance of our solutions.Source: 5th Italian Information Retrieval Workshop, pp. 75–79, University of Roma Tor Vergata, 21-22 January 2014
See at:
ceur-ws.org | CNR ExploRA
2014
Contribution to book
Restricted
Effective Data Access Patterns on Massively Parallel Processors
Capannini G., Baraglia R., Silvestri F., Nardini F. M.The new generation of microprocessors incorporates a huge number of cores on the same chip. This trades single-core performance off for the total amount of work done across multiple threads of execution. Graphics Processing Units (GPUs) are an example of this kind of architectures. The first generation of GPUs has been designed to support a fixed set of rendering functions. Nowa- days, GPUs are becoming easier to program. Therefore, they can be used for applications that have been traditionally handled by CPUs. The reasons of using General Purpose GPU (GPGPUs) in high-performance computations are: raw computing power, good performance per watt, and low costs. How- ever, some important issues limit a wide exploitation of GPGPUs. The main one concerns the heterogeneous and distributed nature of the memory hierar- chy. As a consequence, the speed-up of some applications depends on being able to efficiently access the data so that all cores are able to work at the same time. This chapter discusses the characteristics and the issues of the memory systems of this kind of architectures. We analyze these architectures from a theoretical point by using K-model, a model for capturing their performance constraints. K -model is used to estimate the complexity of a given algorithm defined on this model. This chapter describes how K-model can also be used to design efficient data access patterns for implementing efficient GPU algorithms. To this extent, we use K -model to derive an efficient realization of two popular algorithms, i.e., prefix sum and sorting. By means of reproducible experiments, we validate theoretical results showing that the optimization of an algorithm based on K-model corresponds to an actual optimization in practice.Source: High-Performance Computing on Complex Environments, edited by Emmanuel Jeannot, Julius Zilinskas, pp. 115–134. Hoboken: John Wiley & Sons Inc., 2014
DOI: 10.1002/9781118711897.ch7Metrics:
See at:
doi.org | www.scopus.com | CNR ExploRA
2013
Journal article
Open Access
ATLAAS-P2P: A Two-Layer Architecture for Approximated Search in Peer to Peer
Baraglia R., Dazzi P., Mordacchini M., Ricci L.ATLAAS-P2P is a two-layered peer-to-peer (P2P) architecture for developing systems, providing resource aggregation and approximated discovery in P2P networks. It gives users a flexible and easy means of searching for resources and also benefits resource providers by assisting users to find them. The process of identifying useful resources in a P2P network is highly dependent on query formulation. Users should be able to easily express their needs, and an efficient query resolution mechanism should efficiently find relevant resources and limit the number of messages exchanged. Common techniques for searching resources in P2P systems are based on range queries over a set of attributes. However, the volume of resources in a P2P network may be very large and heterogeneous, and users rarely have the appropriate knowledge about the available resources to allow them to properly formulate their queries. A user may, however, be able to define their "ideal" resource and ask the search system to find resources close to such an entity. Thus, instead of having to specify precise ranges on all attributes, the user simply has to provide an example of what is needed.
This mechanism would simplify the work for users and lead to a more efficient exploitation of the search system. Moreover, it would provide an effective infrastructure for advertising for resource providers, facilitating their discovery by users.Source: ERCIM news 92 (2013).
See at:
ercim-news.ercim.eu | CNR ExploRA
2013
Journal article
Open Access
A peer-to-peer recommender system for self-emerging user communities based on gossip overlays
Baraglia R., Dazzi P., Mordacchini M., Ricci L.Gossip-based peer-to-peer protocols proved to be very efficient for supporting dynamic and complex information exchange among distributed peers. They are useful for building and maintaining the network topology itself as well as to support a pervasive diffusion of the information injected into the network. This is very useful in a world where there is a growing need to access and be aware of many types of distributed resources like Internet pages, shared files, online products, news and information. Finding flexible, scalable and efficient mechanisms addressing this topic is a key issue, even with relevant social and economic aspects. In this paper, we propose the general architecture of a system whose aim is to exploit the collaborative exchange of information between peers in order to build a system able to gather similar users and spread useful suggestions among them.Source: Journal of computer and system sciences (Print) 79 (2013): 291–308. doi:10.1016/j.jcss.2012.05.011
DOI: 10.1016/j.jcss.2012.05.011Project(s): CONTRAIL ,
RECOGNITION ,
S-CUBE Metrics:
See at:
Journal of Computer and System Sciences | www.sciencedirect.com | CNR ExploRA
2013
Contribution to conference
Unknown
First ACM workshop on Optimization techniques for resources management in clouds
Baraglia R., Coppola M., Dazzi P.This workshop has been conceived to be a forum for presentation of research results and experience reports on management and optimization issues related to dependability, scalability, economicity and performance at each level of cloud computing infrastructures. In fact, cloud computing is gaining an increasing degree of popularity and interest both in the industrial and in the scientific community, allowing customers to outsource the management of physical resources by renting in a pay-per-use fashion a variable amount of resources according to their actual needs. Anyhow, despite the results obtained by researchers, advancing in information technology poses many challenges on resource management to hide customers the details of physical resources and to provide a flexible and efficient environment to users. The main focus of the workshop is to share novel Cloud resource optimization solutions that fulfill the needs of heterogeneous applications and environments as well as to identify new directions for future research and development.Project(s): CONTRAIL
See at:
CNR ExploRA
2013
Conference article
Unknown
Il CNR dopo la CEP
Montani C., Andronico P., Raviolo C., Bozzi A., Codenotti B., Meghini C., Sommani M., Tarabella L., Scopigno R., Baraglia R., Perego R.A short history of some of the ICT issues developed in the Institutes of CNR in Pisa since the 60s and that, in the opinion of the authors, had its roots in the CEP (Pisa Electronic Computer).Source: La CEP prima della CEP: storia dell'informatica. Atti, pp. 41–66, Pisa, Italy, 11-12 novembre 2011
See at:
CNR ExploRA
2013
Conference article
Restricted
LearNext: learning to predict tourists movements
Baraglia R., Muntean C. I., Nardini F. M., Silvestri F.In this paper, we tackle the problem of predicting the ``next'' geographical position of a tourist given her history (i.e., the prediction is done accordingly to the tourist's current trail) by means of supervised learning techniques, namely Gradient Boosted Regression Trees and Ranking SVM. The learning is done on the basis of an object space represented by a 68 dimension feature vector, specifically designed for tourism related data. Furthermore, we propose a thorough comparison of several methods that are considered state-of-the-art in touristic recommender and trail prediction systems as well as a strong popularity baseline. Experiments show that the methods we propose outperform important competitors and baselines thus providing strong evidence of the performance of our solutions.Source: CIKM '2013 - 22nd ACM International Conference on Information & Knowledge Management, pp. 751–756, San Francisco, USA, 27 October - 1 November 2013 2013
DOI: 10.1145/2505515.2505656Metrics:
See at:
dl.acm.org | doi.org | www.scopus.com | CNR ExploRA
2013
Contribution to conference
Restricted
Towards GROUP protocol formalization
Baraglia R., Dazzi P., Mordacchini M. Ricci L.Over recent years, we experienced a huge diffusion of internet connected computing devices. As a consequence, this leaded to research for efficient and scalable approaches for managing the burden caused by the highly increased volume of data to be exchanged and processed. Efficient communication protocols are fundamental building blocks for realizing such approaches [1], [2]. Thus, several peer-to-peer protocols have been proposed. Gossip protocols [3]-[7] are a family of peer-to-peer protocols that proved to be well-suited for supporting a scalable and decentralized strategy for peer and data aggregation and diffusion. However, one of the typical limitation of Gossip protocols consists in the selfish behavior adopted by peers in defining their neighborhood and, as a consequence, the topology of the overlay they build. GROUP [8] is a Gossip protocol we conceived to overcome this limitation. It builds explicit defined communities of peers that are identified by their leaders, each one elected in a distributed fashion. This protocol experimentally proved to be efficient and effective with respect to its aim. Anyhow, no analytical study has been realized so far. This work presents a currently ongoing work we are conducting for exploring the properties of GROUP in a more formal way. We conduct this preliminary investigation using a formalization based on Markov chains.Source: P2P 2013 - 2013 IEEE Thirteenth International Conference on Peer-to-Peer Computing, pp. 2–5, 2013 IEEE Thirteenth International Conference on Peer-to-Peer Computing, 2013 IEEE Thirteenth International Conference on Peer-to-Peer Computing
DOI: 10.1109/p2p.2013.6688727Project(s): MIDAS Metrics:
See at:
doi.org | ieeexplore.ieee.org | CNR ExploRA
2013
Journal article
Open Access
A multi-criteria job scheduling framework for large computing farms
R. Baraglia, G. Capannini, P. Dazzi, G. PaganoIn this paper, we propose a new multi-criteria job scheduler for scheduling a continuous stream of batch jobs on large-scale computing farms. Our solution, called Convergent Scheduler, exploits a set of heuristics that drives the scheduler in taking decisions. Each heuristics manages a specific problem constraint, and contributes to compute a value that measures the degree of matching between a job and a machine. Scheduling choices are taken both to meet the Quality of Service requested by the submitted jobs and to optimize the usage of software and hardware resources. In order to validate the scheduler we propose, it has been compared versus two common job scheduling algorithms: Easy and Flexible backfilling. Convergent Scheduler demonstrated to be able to compute good assignments that allow a better exploitation of resources with respect to the other algorithms. Moreover, it has a simple modular structure that makes simple its extension and customization to meet the service goal of an installation.Source: Journal of computer and system sciences (Print) 79 (2013): 230–244. doi:10.1016/j.jcss.2012.05.005
DOI: 10.1016/j.jcss.2012.05.005DOI: 10.1109/cit.2010.69Project(s): ENVRI ,
IMARINE ,
EUBRAZILOPENBIO ,
CONTRAIL Metrics:
See at:
Journal of Computer and System Sciences | doi.org | www.sciencedirect.com | CNR ExploRA
2013
Contribution to book
Unknown
Il CNR dopo la CEP
Claudio Montani, Patrizia Andronico, Claudia Raviolo, Andrea Bozzi, Bruno Codenotti, Carlo Meghini, Marco Sommani, Leonello Tarabella, Roberto Scopigno, Ranieri Baraglia, Raffaele PeregoUna breve [e parziale] storia di alcune tematiche ICT di successo che si sono sviluppate negli Istituti CNR di Pisa a partire dagli anni '60 e che, a giudizio degli autori, rappresentano a buon diritto rami importanti di quell'albero rigoglioso che ha avuto le sue radici nella CEP.Source: La CEP prima della CEP: storia dell'informatica. Divulgazione scientifica e didattica sperimentale. Atti del Convegno, Pisa 11-12 novembre 2011. Pisa: Pisa University Press, 2013
See at:
CNR ExploRA
2013
Contribution to book
Open Access
ORMaCloud 2013 - Chairs' welcome
Baraglia R., Coppola M., Dazzi P.Prefazione agli atti del workshop internazionale ORMaCloud 2013Source: ORMaCloud '13: Proceedings of the first ACM workshop on Optimization techniques for resources management in clouds, edited by Baraglia R.; Coppola M.; Dazzi P.. New York: ACM Press, 2013
See at:
dl.acm.org | ISTI Repository | CNR ExploRA
2012
Journal article
Restricted
Sorting on GPUs for large scale datasets: a thorough comparison
Capannini G., Silvestri F. : Baraglia R.Although sort has been extensively studied in many research works, it still remains a challenge in particular if we consider the implications of novel processor technologies such as manycores (i.e. GPUs, Cell/BE, multicore, etc.). In this paper, we compare different algorithms for sorting integers on stream multiprocessors and we discuss their viability on large datasets (such as those managed by search engines). In order to fully exploit the potentiality of the underlying architecture, we designed an optimized version of sorting network in the K-model, a novel computational model designed to consider all the important features of many-core architectures. According to K-model, our bitonic sorting network mapping improves the three main aspects of many-core architectures, i.e. the processors exploitation, and the on-chip/off-chip memory bandwidth utilization. Furthermore we are able to attain a space complexity of O(1). We experimentally compare our solution with state-of-the-art ones (namely, quick-sort and radix-sort) on GPUs. We also compute the complexity in the K-model for such algorithms. The conducted evaluation highlight that our bitonic sorting network is faster than quick-sort and slightly slower than radix, yet being an in-place solution it consumes less memory than both algorithms.Source: Information processing & management 48 (2012): 903–917. doi:10.1016/j.ipm.2010.11.010
DOI: 10.1016/j.ipm.2010.11.010Metrics:
See at:
Information Processing & Management | www.sciencedirect.com | CNR ExploRA
2012
Conference article
Restricted
RecTour: a recommender system for tourists
Baraglia R., Frattari C., Muntean C. I., Nardini F. M., Silvestri F.This paper presents a recommender system that provides personalized information about locations of potential interest to a tourist. The system generates suggestions, consisting of touristic places, according to the current position and history data describing the tourist movements. For the selection of tourist sites, the system uses a set of points of interest a priori identified. We evaluate our system on two datasets: a real and a synthetic one, both storing trajectories describing previous movements of tourists. The proposed solution has high applicability and the results show that the solution is both efficient and viable.Source: International Workshop on Tourism Facilities, Macau, China, 4 December 2012
DOI: 10.1109/wi-iat.2012.88Metrics:
See at:
doi.org | ieeexplore.ieee.org | CNR ExploRA
2012
Conference article
Restricted
A multi-criteria class-based job scheduler for large computing farms.
Baraglia R., Dazzi P., Ferrini R.In this paper we propose a new multi-criteria class-based job scheduler able to dynamically schedule a stream of batch jobs on large-scale computing farms. It is driven by several configuration parameters allowing the scheduler customization with respect to the goals of an installation. The proposed scheduling policies allow to maximize the resource usage and to guarantee the ap- plications QoS requirements. The proposed solution has been evaluated by simulations using different streams of synthetically generated jobs. To analyze the quality of our solution we propose a new methodology to estimate whether at a given time the resources in the system are really sufficient to meet the service level requested by the submitted jobs. Moreover, the proposed solution was also evaluated comparing it with the Backfilling and Flexible backfilling algorithms. Our scheduler demonstrated to be able to carry out good scheduling choices.Source: International Conference on Parallel and Distributed Processing Techniques and Applications, pp. 289–295, Las Vegas, USA, 16-19 July 2012
Project(s): CONTRAIL
See at:
world-comp.org | CNR ExploRA
2012
Conference article
Open Access
A trajectory-based recommender system for tourism.
Baraglia R., Frattari C., Muntean C. I., Nardini F. M., Silvestri F.Recommendation systems provide focused information to users on a set of objects belonging to a specific domain. The proposed recommender system provides personalized suggestions about touristic points of interest. The system generates recommendations, consisting of touristic places, according to the current position of a tourist and previously collected data describing tourist movements in a touristic location/city. The touristic sites correspond to a set of points of interest identified a priori. We propose several metrics to evaluate both the spatial coverage of the dataset and the quality of recommendations produced. We assess our system on two datasets: a real and a synthetic one. Results show that our solution is a viable one.Source: Active Media Technology. 8th International Conference, pp. 196–205, Macau, China, 4-7 December 2012
DOI: 10.1007/978-3-642-35236-2_20Metrics:
See at:
hpc.isti.cnr.it | doi.org | link.springer.com | CNR ExploRA
2012
Journal article
Open Access
A Peer-to-Peer recommender system for self-emerging user communities based on Gossip overlays
Baraggia R., Dazzi P., Mordacchini M., Ricci L.Gossip-based peer-to-peer protocols proved to be very efficient for supporting dynamic and complex information exchange among distributed peers. They are useful for building and maintaining the network topology itself as well as to support a pervasive diffusion of the information injected into the network. This is very useful in a world where there is a growing need to access and be aware of many types of distributed resources like Internet pages, shared files, online products, news and information. Finding flexible, scalable and efficient mechanisms addressing this topic is a key issue, even with relevant social and economic aspects. In this paper, we propose the general architecture of a system whose aim is to exploit the collaborative exchange of information between peers in order to build a system able to gather similar users and spread useful suggestions among them.Source: Journal of computer and system sciences (Print) 79 (2012): 291–308. doi:10.1016/j.jcss.2012.05.011
DOI: 10.1016/j.jcss.2012.05.011Project(s): CONTRAIL ,
RECOGNITION ,
S-CUBE Metrics:
See at:
Journal of Computer and System Sciences | www.sciencedirect.com | CNR ExploRA