Research Interests:
ABSTRACT Heterogeneous computing is a special form of parallel and distributed computing where computations are performed using a single autonomous computer operating in both SIMD and MIMD modes, or using a number of connected autonomous... more
ABSTRACT Heterogeneous computing is a special form of parallel and distributed computing where computations are performed using a single autonomous computer operating in both SIMD and MIMD modes, or using a number of connected autonomous computers. In multimode system heterogeneous computing, tasks can be executed in both SIMD and MIMD simultaneously. In this paper, we present PQE HPF, a High Performance Fortran (HPF) based programming library which allows one to exploit the MIMD and SIMD capabilities offered by PQE-1, a multimode parallel architecture. Two different implementations of a well-known application, using HPF and PQE HPF respectively, were used to evaluate the overheads introduced over the machine's runtime system. Preliminary tests, conducted by running the case study application on the first PQE-1 prototype, show good results and encourage us to dedicate more effort to implement real production parallel codes on a similar architecture
Research Interests:
Although high{level data-parallel languages, such as High Performance Fortran (HPF), provide an essential tool to deal with various numerical problems, a more general approach is often needed that integrates both task and data... more
Although high{level data-parallel languages, such as High Performance Fortran (HPF), provide an essential tool to deal with various numerical problems, a more general approach is often needed that integrates both task and data parallelism. In the context of a project aimed at building a new hybrid architecture composed of both MIMD and SIMD computing resources, we have developed a library which allow programmers to exploit the SIMD computational elements from within their HPF programs. Task parallelism can be thus achieved by concurrently running MIMD and SIMD parts of the same application. The overheads introduced over the machine's run-time system are found to be negligible. Preliminar tests conducted by running a physics application on the rst prototype of the hybrid architecture show encouraging results. 1 Introduction Data{parallel languages such as High Performance Fortran (HPF) provide a powerful tool for many computationally-hard problems. They let programmers to ...
The length of each indexed document is needed by most common text retrieval scoring functions to rank it with respect to the current query. For efficiency purposes information retrieval systems maintain this information in the main... more
The length of each indexed document is needed by most common text retrieval scoring functions to rank it with respect to the current query. For efficiency purposes information retrieval systems maintain this information in the main memory. This paper proposes a novel strategy to encode the length of each document directly in the document identifier, thus reducing main memory demand. The technique is based on a simple document identifier assignment method and a function allowing the approximate length of each indexed ...
Research Interests:
Research Interests:
Abstract. Genetic Algorithms (GAs) [4] are stochastic optimization heuristics in which searches in solution space are carried out by imi-taring the population genetics stated in Darwin's theory of evolution. In order to apply... more
Abstract. Genetic Algorithms (GAs) [4] are stochastic optimization heuristics in which searches in solution space are carried out by imi-taring the population genetics stated in Darwin's theory of evolution. In order to apply GAs to a problem, a genetic representation of each ...
Research Interests:
Abstract We study the problem of diversifying search results by exploiting the precious knowledge stored in query logs. When an ambiguous or faceted query is submitted to a search engine, the user has often to reformulate it in order to... more
Abstract We study the problem of diversifying search results by exploiting the precious knowledge stored in query logs. When an ambiguous or faceted query is submitted to a search engine, the user has often to reformulate it in order to better satisfy her information need by improving the perceived precision of results returned. Our proposal exploits the presence of dierent\ specializations" of queries in query logs to detect the submission of ambiguous/faceted queries, and manage them by diversifying the search results returned ...
... Deliverable type: Report Classification: R Work package and task: WP4, T4.3 Responsibility: CNR Editor: Claudio Lucchese Contributors: Claudio Lucchese, Raffaele Perego, Fausto Rabitti, Fabrizio Falchi. Internal Reviewer: Maristella... more
... Deliverable type: Report Classification: R Work package and task: WP4, T4.3 Responsibility: CNR Editor: Claudio Lucchese Contributors: Claudio Lucchese, Raffaele Perego, Fausto Rabitti, Fabrizio Falchi. Internal Reviewer: Maristella Agosti Deliverable D4.4 ...
Research Interests:
Research Interests:
Research Interests:
This paper presents a new framework aimed at compile-time determination of satisfactory sub-optimal solutions to the mapping problem onto modern massively parallel computing systems. The approach incorporates realistic assumptions on the... more
This paper presents a new framework aimed at compile-time determination of satisfactory sub-optimal solutions to the mapping problem onto modern massively parallel computing systems. The approach incorporates realistic assumptions on the models both for parallel ...
Research Interests:
The recommendation of queries, known as query suggestion, is a common practice on major Web Search Engines. It aims to help users to find the information they are looking for, and is usually based on the knowledge learned from past... more
The recommendation of queries, known as query suggestion, is a common practice on major Web Search Engines. It aims to help users to find the information they are looking for, and is usually based on the knowledge learned from past interactions with the search engine. In this paper we propose a new model for query suggestion, the Search Shortcut Problem,
Research Interests:
Research Interests:
ABSTRACT We propose TripBuilder, an user-friendly and interactive system for planning a time-budgeted sightseeing tour of a city on the basis of the points of interest and the patterns of movements of tourists mined from user-contributed... more
ABSTRACT We propose TripBuilder, an user-friendly and interactive system for planning a time-budgeted sightseeing tour of a city on the basis of the points of interest and the patterns of movements of tourists mined from user-contributed data. The knowledge needed to build the recommendation model is entirely extracted in an unsupervised way from two popular collaborative platforms: Wikipedia 1 and Flickr 2. Trip-Builder interacts with the user by means of a friendly Web interface 3 that allows her to easily specify personal interests and time budget. The sightseeing tour proposed can be then explored and modified. We present the main components composing the system.
Summary. The calculation of rate constant values of the H +//2 reaction for an extended range of excited vibrational states of the diatomic molecule and temperatures is relevant to the modeling of H-sources. To investigate the effect of... more
Summary. The calculation of rate constant values of the H +//2 reaction for an extended range of excited vibrational states of the diatomic molecule and temperatures is relevant to the modeling of H-sources. To investigate the effect of isotopic substitutions on the efficiency of ...
Research Interests:
The suitability of massively parallel architectures for carrying out efficient calculations of quasiclassical rate constants for atom-diatom reactive processes has been investigated. Problems related to the parallel structuring of the... more
The suitability of massively parallel architectures for carrying out efficient calculations of quasiclassical rate constants for atom-diatom reactive processes has been investigated. Problems related to the parallel structuring of the computational procedure, fixed and scaled speed-ups, efficiency factors and their dependence upon the size of the problem, and the number of processors are discussed.
Research Interests:
Research Interests:
Research Interests:
Abstract The paper deals with the parallelization of Delaunay triangulation algorithms, giving more emphasis to pratical issues and implementation than to theoretical complexity. Two parallel implementations are presented. The first one... more
Abstract The paper deals with the parallelization of Delaunay triangulation algorithms, giving more emphasis to pratical issues and implementation than to theoretical complexity. Two parallel implementations are presented. The first one is built on De Wall, an Ed triangulator based ...
Research Interests:
D4.5 Implementation of software prototype for P2P indexing and Collaborative Crawling, Dec 2008 ... Project acronym: SAPIR Project full title: Search on Audio-visual content using Peer-to-Peer Information Retrieval Contract no.: 45128... more
D4.5 Implementation of software prototype for P2P indexing and Collaborative Crawling, Dec 2008 ... Project acronym: SAPIR Project full title: Search on Audio-visual content using Peer-to-Peer Information Retrieval Contract no.: 45128 ... Deliverable type: Prototype ...
... Deliverable type: Report Classification: R Work package and task: WP4, T4.3 Responsibility: CNR Editor: Claudio Lucchese Contributors: Claudio Lucchese, Raffaele Perego, Fausto Rabitti, Fabrizio Falchi. Internal Reviewer: Maristella... more
... Deliverable type: Report Classification: R Work package and task: WP4, T4.3 Responsibility: CNR Editor: Claudio Lucchese Contributors: Claudio Lucchese, Raffaele Perego, Fausto Rabitti, Fabrizio Falchi. Internal Reviewer: Maristella Agosti Deliverable D4.4 ...
Research Interests:
EXECUTIVE SUMMARY This report presents the activities conducted within task T4. 3 of the SAPIR project. It discusses the state-of-the-art of caching techniques in peer-to-peer (P2P) information systems. Caching is widely used in... more
EXECUTIVE SUMMARY This report presents the activities conducted within task T4. 3 of the SAPIR project. It discusses the state-of-the-art of caching techniques in peer-to-peer (P2P) information systems. Caching is widely used in information systems to achieve better performance and high throughput. When moving to large scale distributed information systems, exploiting caching and replication becomes a must in order to overcome network costs and improve response times. Since caching in P2P information retrieval systems is ...
Although HPF allows programmers to express data-parallel computations in a portable, high-level way, it is widely accepted that many important parallel applications cannot be efficiently implemented following a pure data-parallel... more
Although HPF allows programmers to express data-parallel computations in a portable, high-level way, it is widely accepted that many important parallel applications cannot be efficiently implemented following a pure data-parallel paradigm. For these applications, rather than having a single data-parallel program, it is more profitable to subdivide the whole computation into several data-parallel pieces, where the various pieces run concurrently and co-operate, thus exploiting task parallelism. This paper discusses the ...
Research Interests:
This paper describes the architecture of MOSE (My Own Search Engine), a scalableparallel and distributed engine for searching the web. MOSE was specicallydesigned to eciently exploit aordable parallel architectures, such as clusters... more
This paper describes the architecture of MOSE (My Own Search Engine), a scalableparallel and distributed engine for searching the web. MOSE was specicallydesigned to eciently exploit aordable parallel architectures, such as clusters ofworkstations. Its modular and scalable architecture can be easily adjusted to ful-ll the bandwidth requirements of the application at hand. Both task-paralleland data-parallel approaches are exploited within MOSE in order to increasethe throughput and eciently use communication, storing and ...
Abstract. In this paper, we deal with efficiency of the diversification of results returned by Web Search Engines (WSEs). We extend a search architecture based on additive Machine Learned Ranking (MLR) systems with a new module computing... more
Abstract. In this paper, we deal with efficiency of the diversification of results returned by Web Search Engines (WSEs). We extend a search architecture based on additive Machine Learned Ranking (MLR) systems with a new module computing the diversity score of each retrieved document. Our proposed solution is designed to be used with other techniques,(eg early termination of rank computation, etc.). Furthermore, we use an efficient state-of-the-art diversification approach based on knowledge extracted from query logs, and prove that it ...
Europeana is a strategic project funded by the European Commission with the goal of making Europe's cultural and scientific heritage accessible to the public. ASSETS is a two-year Best Practice Network co-funded by the CIP PSP... more
Europeana is a strategic project funded by the European Commission with the goal of making Europe's cultural and scientific heritage accessible to the public. ASSETS is a two-year Best Practice Network co-funded by the CIP PSP Programme to improve performance, accessibility and usability of the Europeana search engine. Here we present a characterization of the Europeana logs by showing statistics on common behavioural patterns of the Europeana users.
Research Interests:
Research Interests:
Welcome to the Euro-Par 2001 Topic 03 on Scheduling and Load Balancing. Scheduling and load balancing are key areas in the quest for performance in parallel and distributed applications. Relevant techniques can be provided either at the... more
Welcome to the Euro-Par 2001 Topic 03 on Scheduling and Load Balancing. Scheduling and load balancing are key areas in the quest for performance in parallel and distributed applications. Relevant techniques can be provided either at the applicatio level, or at the system level, and both scenarios are of interest for this topic.
Research Interests:
Research Interests:
Research Interests:
Research Interests:
UK PubMed Central (UKPMC) is an archive of life sciences journal literature.
Due to the huge increase in the number and dimension ofavailable databases, e# cient solutions for counting frequent sets are nowadaysvery important within the Data Mining community. Several sequentialand parallel algorithms were... more
Due to the huge increase in the number and dimension ofavailable databases, e# cient solutions for counting frequent sets are nowadaysvery important within the Data Mining community. Several sequentialand parallel algorithms were proposed, which in many cases exhibitexcellent scalability. In this paper we present ParDCI, a distributed andmultithreaded algorithm for counting the occurrences of frequent setswithin transactional databases. ParDCI is a parallel version of DCI (DirectCount & Intersect), a multi-strategy algorithm ...
This article introduces an architecture for a document-partitioned search engine, based on a novel approach combining collection selection and load balancing, called load-driven routing. By exploiting the query-vector document model, and... more
This article introduces an architecture for a document-partitioned search engine, based on a novel approach combining collection selection and load balancing, called load-driven routing. By exploiting the query-vector document model, and the incremental caching technique, our architecture can compute very high quality results for any query, with only a fraction of the computational load used in a typical document-partitioned architecture. By trading off a small fraction of the results, our technique allows us to strongly reduce the computing pressure to a search engine back-end; we are able to retrieve more than 2/3 of the top-5 results for a given query with only 10% the computing load needed by a configuration where the query is processed by each index partition. Alternatively, we can slightly increase the load up to 25% to improve precision and get more than 80% of the top-5 results. In fact, the flexibility of our system allows a wide range of different conf...