Versioning

Cloud computing offers a powerful abstraction that provides a scalable, virtualized infrastructure as a service where the complexity of fine-grained resource management is hidden from the end-user. Running data analytics applications in the cloud on extremely large data sets is ...

Bancos de dados e documentos sao comumente mantidos em separado nas organizacoes, controlados por Sistemas Gerenciadores de Bancos de Dados (SGBDs) e Sistemas de Recuperacao de Informacao (SRIs), respectivamente. Essa separacao tem ligacao com a natureza dos dados manipulados: estruturados, no primeiro caso; nao estruturados, no segundo. Enquanto os SGBDs processam consultas exatas a bancos de dados, os SRIs recuperam documentos com base em buscas por palavras-chave, que sao inerentemente imprecisas. Apesar disso, a integracao desses sistemas pode resultar em grandes ganhos ao usuario, uma vez que, numa mesma organizacao, bancos de dados e documentos frequentemente se referem a entidades comuns. Uma das possibilidades de integracao e a recuperacao de documentos associados a uma dada consulta a banco de dados. Por exemplo, considerando a consulta "Quais os clientes com contratos acima de X reais?", como recuperar documentos que possam estar associados a esta consulta, como ...

The present paper aims to shed light on the experiences of racism faced by second-generation black immigrants in Canada and explore their quest for identity in the context of a multicultural Canada. David Chariandy's novel Brother, depicts the discrimination endured by black ethnic minorities, along with the intercultural negotiation in the formation of a hybrid identity influenced by subversive elements of resistance such as Hip-Hop culture. Chariandy adds to the growing body of counter-narratives that challenge preconceived notions surrounding Canadian multicultural policies, exposing the lingering racism that the nation seeks to conceal.

In unstructured distributed P2P systems there is no logical structure to control the peers coming and leaving the network, which can occur anytime due to mobility. Thus, data exchange with consistence, and data availability are very important. To favor high data availability and to acquire performance gains in dynamic behavior environments, P2P systems use data replication. However, current replication data models have a single failure point, or require complex conciliation algorithms to acquire consistency. This work introduces a replication data model that simplifies the consistency mechanism.

This paper considers a two-stage development problem for information goods with costless quality degradation. In our model, a seller of information goods faces customers that are heterogeneous with regard to both the marginal willingness to pay for quality and the outside opportunity. In the development stage, the seller determines the quality limit of the product. In the second stage, the seller’s problem is to design the price schedule corresponding to different quality levels, taking into account production and distribution costs. We show that versioning is optimal for the seller when customers have multiple outside options or, more generally, convex reservation utilities. In addition, we show that in the optimal solution the seller discards both low-end and high-end customers. Among those that are served, the seller offers a continuum of (inferior) versions to customers with relatively low willingness to pay, and extracts full information rent from each of them. A common version...

It is commonly believed that piracy of information goods leads to lower profits, which translate to lower incentives to invest in innovation, and eventually to lower quality products. Manufacturers, policy-makers, and researchers, all claim that inadequate piracy enforcement efforts translate to lower investments in product development. However, we find many practical examples that contradict this claim. Therefore, in order to examine this claim more carefully, we develop a rigorous economic model of the manufacturer's quality decision problem in the presence of piracy. We consider a monopolist, who does not have any marginal costs, but has a product development cost that is quadratic in the quality level produced. The monopolist faces a consumer market that is heterogeneous in its preference for quality, and offers a quality level that maximizes its profit. We also allow for the possibility that the manufacturer may use versioning to counter piracy. We unexpectedly find that, in certain situations, a lower piracy enforcement increases the monopolist's incentive to invest in quality. We explain the reasons and welfare implications of our findings.

With sinking storage costs, it becomes more and more feasible, and popular, to retain past versions of documents and data. While undoing changes is worthy, this becomes even more valuable if the data is queryable. Nowadays, there are two widespread version control paradigms: document versioning (SVN, git, etc.) and versioned databases. The former handles any kind of document, even binary, but only sees lines of text, so that the query capability is limited. The latter provide fine-grained temporal query capabilities on highly structured data-but storing everything in a relational database is not desirable. The goal of this paper is to provide a unified framework for efficiently versioning, querying and updating not only data and documents, but also, inbetween, any kind of semi-structured information, like XML. We start with the XQuery programming language and meticulously extend its data model, its syntax and its processing model to make it seamlessly time-aware. We provide data structures and algorithms for the efficient implementation of such a versioning system. Finally, we show that there is no significant performance loss for traditional queries when enriching an existing engine with versioning capabilities.

The maintenance of materialized views in large-scale environments composed of numerous information sources (ISs), such as in the WWW, is complicated by ISs not only continuously modifying their contents but also their capabilities (schemas and query interfaces). With current view technology, views become undefined when ISs change their capabilities. Our Evolvable View Environment (EVE) project addresses this new problem of evolving views under IS capabilities changes, which we coin view synchronization problem. Key principles of EVE include a userspecified preference model for view evolution (Evolvable-SQL (E-SQL)) and a Model for Information Source Descriptions (MISD). In this paper, we first present a formal characterization of correctness of view synchronization using containment constraints defined in MISD. Then, we give a novel view synchronization algorithm for view rewriting exploiting general containment constraints between the tobe-replaced relation and its replacement.

The trajectory data warehouse (TDW) view definitions are constructed from heterogeneous mobile information sources schema that are more and more independent. In fact, they frequently change their content due to perpetual transactions (data changes) and may change their structure due to continual users' requirements evolving. Managing appropriately the view definition synchronization is a necessity since, the TDW is considered as the core component of the modern decision support systems. The aim of this paper is to propose a mobile agent view synchronization system to uphold a trajectory data warehouse under schema changes.

File-sharing semantics is used by the file systems for sharing data among concurrent client processes in a consistent manner. Session semantics is a widely used file-sharing semantics in Distributed File Systems (DFSs). The main disadvantage of session semantics is that writes to an open file are visible to the concurrent client processes only during their next session. Recently, "linearizability semantics" was introduced in BlobSeer DFS, in which a Read Client Process (RCP) can read only a previous version of a binary large object (blob), while update operations are carried out on that blob in a concurrent manner. In this paper, we propose a new type of file-sharing semantics, namely "speculative semantics", which permits writes to an open file to be visible to other concurrent processes provided that data consistency is not affected. In addition, we propose a new read algorithm for DFSs based on speculative semantics and a new performance measurement metric called Currency. The experimental results obtained using BlobSeer DFS indicate that the proposed read algorithm performs better than the existing read algorithm of BlobSeer DFS.

Hadoop Distributed File System (HDFS) is the core component of Apache Hadoop project. In HDFS, the computation is carried out in the nodes where relevant data is stored. Hadoop also implemented a parallel computational paradigm named as Map-Reduce. In this paper, we have measured the performance of read and write operations in HDFS by considering small and large files. For performance evaluation, we have used a Hadoop cluster with five nodes. The results indicate that HDFS performs well for the files with the size greater than the default block size and performs poorly for the files with the size less than the default block size.

Background: Provenance is a critical ingredient for establishing trust of published scientific content. This is true whether we are considering a data set, a computational workflow, a peer-reviewed publication or a simple scientific claim with supportive evidence. Existing vocabularies such as Dublin Core Terms (DC Terms) and the W3C Provenance Ontology (PROV-O) are domain-independent and general-purpose and they allow and encourage for extensions to cover more specific needs. In particular, to track authoring and versioning information of web resources, PROV-O provides a basic methodology but not any specific classes and properties for identifying or distinguishing between the various roles assumed by agents manipulating digital artifacts, such as author, contributor and curator. Results: We present the Provenance, Authoring and Versioning ontology (PAV, namespace http://purl.org/pav/): a lightweight ontology for capturing "just enough" descriptions essential for tracking the provenance, authoring and versioning of web resources. We argue that such descriptions are essential for digital scientific content. PAV distinguishes between contributors, authors and curators of content and creators of representations in addition to the provenance of originating resources that have been accessed, transformed and consumed. We explore five projects (and communities) that have adopted PAV illustrating their usage through concrete examples. Moreover, we present mappings that show how PAV extends the W3C PROV-O ontology to support broader interoperability. Method: The initial design of the PAV ontology was driven by requirements from the AlzSWAN project with further requirements incorporated later from other projects detailed in this paper. The authors strived to keep PAV lightweight and compact by including only those terms that have demonstrated to be pragmatically useful in existing applications, and by recommending terms from existing ontologies when plausible. Discussion: We analyze and compare PAV with related approaches, namely Provenance Vocabulary (PRV), DC Terms and BIBFRAME. We identify similarities and analyze differences between those vocabularies and PAV, outlining strengths and weaknesses of our proposed model. We specify SKOS mappings that align PAV with DC Terms. We conclude the paper with general remarks on the applicability of PAV.

Tomando como referência o modelo editorial que está a ser desenvolvido para a poesia de Pedro Homem de Mello (1904- 1984), este trabalho procura refletir sobre o problema da atribuição crítica de autoridade, nos casos em que se documentam situações de revisão colaborativa e apropriação cultural. Em linha com as posições teóricas de James Thorpe (1972) e Siegfried Scheibe (1995), entre outros, defende-se que, do ponto de vista editorial, todos os testemunhos sancionados pelo autor estão em pé de igualdade, e por isso, em vez de fixar um texto, apresentando-o como a lição mais autêntica, a solução adotada passará por um tratamento não-hierarquizado de múltiplos estágios textuais. Este procedimento, designado de versioning ou multiple texts, beneficia-se especialmente das virtualidades do hipertexto eletrônico, para conciliar tipos variáveis de autoridade e percursos alternativos de leitura.

Immeasurable thanks goes to the Almighty God for granting me the grace and direction to complete this thesis. My gratitude goes to my parents Mr. and Mrs. Daniel Ani and my siblings, my family members, friends and my dear Sarah Garba Ladidi, who all encouraged and prayed for me throughout the period of my studies. I love you all. My profound gratitude goes to my supervisor Dr. Gregory Epiphaniou for the knowledge he had imparted to me as my lecturer and overseer throughout this thesis. To all my tutors, I am ever grateful for your time, encouragement and direction throughout the course of my project and study in University of Bedfordshire. Given the opportunity, working with you again remains my earnest desire. Do accept my appreciation. My appreciation also goes to my sponsor; National Information Technology Development Agency (NITDA), ably led by Prof. Cleopas Angaye for the scholarship opportunity it had offered me and the current support that ensued. It was a rare privilege for me. I would also like to thank my friends

User-Defined Functions (UDF) allow application programmers to specify analysis operations on data, while leaving the data management tasks to the system. This general approach enables numerous custom analysis functions and is at the heart of the modern Big Data systems. Even though the UDF mechanism can theoretically support arbitrary operations, a wide variety of common operationssuch as computing the moving average of a time series, the vorticity of a fluid flow, etc.,-are hard to express and slow to execute. Since these operations are traditionally performed on multi-dimensional arrays, we propose to extend the expressiveness of structural locality for supporting UDF operations on arrays. We further propose an in situ UDF mechanism, called ArrayUDF, to implement the structural locality. ArrayUDF allows users to define computations on adjacent array cells without the use of join operations and executes the UDF directly on arrays stored in data files without requiring to load their content into a data management system. Additionally, we present a thorough theoretical analysis of the data access cost to exploit the structural locality, which enables ArrayUDF to automatically select the best array partitioning strategy for a given UDF operation. In a series of performance evaluations on large scientific datasets, we have observed that-using the generic UDF interface-ArrayUDF consistently outperforms Spark, SciDB, and RasDaMan.

The ubiquity of Big Data has greatly influenced the direction and the development of storage technologies. To meet the needs of storing and analyzing Big Data, researchers and administrators have turned to parallel and distributed storage and compute architectures in both industry and science. While the problems of securely and consitently storing and accessing data in large parallel and distributed file systems have been addressed in both the research and production systems, the indexing and search through large unstructured data and metadata has largely been overlooked. According to the International Data Corporation, more than 90% of data found in the digital universe is unstructured, emphasizing the importance of developing efficient solutions for querying distributed data. This paper proposes a novel indexing solution, called FusionDex, that significantly improves the performance of querying across distributed file systems. FusionDex leverages state-of-the-art, open-source indexing modules as its building blocks to deliver an integrated system for enabling efficient user-specified queries over distributed and unstructured data. FusionDex has been evaluated on a cluster of 64 nodes, and results show that it outperforms existing tools (in some cases by orders of magnitude), such as Hadoop Grep and Cloudera Search.

These last years, the amount of data generated by information systems has exploded. It is not only the quantities of information that are now estimated in Exabyte, but also the variety of these data which is more and more structurally heterogeneous and the velocity of generation of these data which can be compared in many cases to endless flows. Now days, Big Data science offers many opportunities to analyze and explore these quantities of data. Therefore, we can collect and parse data, make many distributed operations, aggregate results, make reports and synthesis. To allow all these operations, Big Data Science relies on the use of "Distributed File Systems (DFS)" technologies to store data more efficiently. Distributed File Systems were designed to address a set of technological challenges like consistency and availability of data, scalability of environments, competitive access to data or even more the cost of their maintenance and extension. In this paper, we attempt to highlight some of these systems. Some are proprietary such as Google GFS and IBM GPFS, and others are open source such as HDFS, Blobseer and AFS. Our goal is to make a comparative analysis of the main technological bricks that often form the backbone of any DFS system.

Modern day systems are facing an avalanche of data, and they are being forced to handle more and more data intensive use cases. These data comes in many forms and shapes: Sensors (RFID, Near Field Communication, Weather Sensors), transaction logs, Web, social networks etc. As an example, weather sensors across the world generate a large amount of data throughout the year. Handling these and similar data require scalable, efficient, reliable and very large storages with support for efficient metadata based searching. This paper present Mahasen, a highly scalable storage for high volume data intensive applications built on top of a peer-to-peer layer. In addition to scalable storage, Mahasen also supports efficient searching, built on top of the Distributed Hash table (DHT)

Web repositories are large scale warehouses of data downloaded from the Web, needed by applications that summarize that data to produce results that help people use information. Time is a central dimension in Web data, because the Web is continuously changing and it is impossible to get a snapshot of a large portion of the Web instantaneously. Developers of applications that manage Web data usually distribute the operations performed by their applications over several processing nodes, to scale-up to the amount of data that may be processed. Versus is a model for a repository providing time oriented distributed Web data management. Time is managed by versioning the objects saved in the repository. Distribution is managed by using a hierarchy of workspaces. Distributed threads work on data stored in the lower level workspaces, and save it by checking-in that data into the workspace in the next upper level. Versus applications can specify the granularity of the distribution and the conflict resolution policies they want to implement. This allows a great control over the repository, increasing the number and type of applications it is suitable to support. The Versus model was embodied in a prototype that is being used to build applications for managing data collected from the Web.

Atualmente, um massivo volume de dados tem sido produzido pelos mais variados tipos de fontes de dados. A facilidade de acesso a esses dados apresenta novas oportunidades, no entanto, escolher quais fontes de dados são mais adequadas para um determinado uso tornou-se um desafio. A literatura oferece diversos trabalhos que abordam a avaliação da qualidade em fontes de dados como meio para solucionar esse desafio, entretanto, poucos trabalhos consideram o aspecto dinâmico das fontes. Neste trabalho, abordamos o problema de avaliação da qualidade em fontes de dados dinâmicas. Além disso, propomos a criação de um Perfil de Qualidade, que consiste de um conjunto de metadados que oferece dados sobre a qualidade de uma fonte e que poderá ser utilizado para facilitar o processo de seleção de fontes de dados. Os experimentos realizados demonstraram que a estratégia de avaliação da qualidade proposta produz resultados satisfatórios.

Gerenciamento de emergência é uma tarefa complexa pois envolve comunicação e colaboração entre variadas organizações e seus sistemas. Integração de dados e interoperabilidade de sistemas estão entre os maiores desafios nesta área. Como solução, este artigo apresenta uma proposta de integração de dados para Sistemas de Resposta à Emergências (SRE) baseado em uma suíte de ontologias criadas a partir do padrão EDXL (Emergency Data Exchange Language). Os resultados indicam que EDXL é adequado ao cenário local e que as ontologias podem ser aplicadas como interlíngua entre os SREs.

As data volumes increase at a high speed in more and more application fields of science, engineering, information services, etc., the challenges posed by data-intensive computing gain an increasing importance. The emergence of highly scalable infrastructures, e.g. for cloud computing and for petascale computing and beyond introduces additional issues for which scalable data management becomes an immediate need. This paper brings several contributions. First, it proposes a set of principles for designing highly scalable distributed storage systems that are optimized for heavy data access concurrency. In particular, we highlight the potentially large benefits of using versioning in this context. Second, based on these principles, we propose a set of versioning algorithms, both for data and metadata, that enable a high throughput under concurrency. Finally, we implement and evaluate these algorithms in the BlobSeer prototype, that we integrate as a storage backend in the Hadoop MapReduce framework. We perform extensive microbenchmarks as well as experiments with real MapReduce applications: they demonstrate that applying the principles defended in our approach brings substantial benefits to data intensive applications.

The use of ontology is widely spread among software engineering groups as a way to represent, structure, share and reuse knowledge. As projects progress, the ontological understanding of the domain may change, evolve. New domain concepts may emerge and existing ones may disappear or be updated, causing changes in the ontology. Taking this into account, the management of the ontology evolution/life cycle should be addressed by the ones that adopt them as a way of representing knowledge. This paper provides an analysis of tools related to ontology evolution management focusing on the ones able to identify elementary and complex changes (by grouping elementary changes) that help on externalizing the intention of the user. The main contribution of this analysis is to present the state of the art of tools related to ontology evolution by identifying their strengths and limitations. The results are particularly useful to ontology designers who need to choose a tool to help them to inspect, understand and manage ontology evolution. Besides, we point out several research issues to be addressed in different kinds of research initiatives.

With the emergence of Cloud Computing, the amount of data generated in different fields such as physics, medical, social networks, etc. is growing exponentially. This increase in the volume of data and their large scale make the problem of their processing more complex. Actually, the current datasets are very different in nature, ranging from small to very large, from structured to unstructured, and from largely complete to noisy and incomplete. In addition, these datasets evolve over time, often at very rapid rates. If we consider the characteristics of these datasets, traditional data management systems are not adapted to support them. For example, Relational Database Management Systems (RDMS) manage only databases where data conforms to a schema. However, current databases contain a mix of structured and less or no structured data. Furthermore, relational systems lack support for version management that is very important in a data management system. As data management system dedicated to large-scale datasets, we consider the BlobSeer system. It is a concurrencyoptimized data management system for data-intensive distributed applications. BlobSeer is adapted for target applications that handle massive unstructured data in the context of large-scale distributed environments. It uses the concept of versioning for concurrent manipulation of large binary objects in order to exploit efficiently access to data. To reach this objective, BlobSeer uses a versioning manager to generate a new snapshot version of a BLOB every time it is written or appended by a client. But if the number of BLOBs created or the primitives (writing, appending or reading) increase and are managed by a single version manager, then we have a performance bottleneck and a version manager overload. To avoid the bottleneck of the version manager, we propose a multi-version managers, such that each version manager maintains a subset of BLOBS.

In this era of developing technologies, one of the most promising is cloud computing that has been functioning since years and used by individuals and large enterprises to provide different kind of services to the world. Cloud computing is attractive to business owners as it eliminates the requirement for users to plan ahead for pro-visioning hardware and allows enterprises to start from the small scale and increase resources only when there is a rise in service demand. One of the trends in cloud computing is the use of decentralized cloud, where resources are shared between multiple servers instead of one central server. The use of decentralized cloud can reduce task complexity and work efficiency per server by utilizing small amount of resources from various servers. With this emerging trend in decentralized cloud, various methodologies were introduced to implement it. One of the most important methodology that plays a vital role in decentralized cloud is Load distribution. Load distribution can be achieved in various ways and hence different approaches can lead to better ways to distribute load in cloud servers. So by researching various methodologies used in cloud computing, a similar but new load distribution algorithm can be implemented which will be based on current statistics of the server nodes by comparing statistics the resultant best node can be chosen to dispatch the file which is to be uploaded.

Deduplicação de registros (DR) tem como objetivo identificar instâncias que representam a mesma entidade do mundo real em repositórios de dados. No ambiente governamental, o processo de DR facilita a identificação de irregularidades e reduz o consumo de recursos computacionais em tarefas de integração de dados. Nesse contexto, propomos neste artigo uma plataforma escalável, efetivo e eficiente para integração de grandes repositórios de dados (i.e., com grandes volumes de dados, da ordem de milhões de registros) para unificar entidades duplicadas oriundas de múltiplas e diferentes fontes. Nossos resultados experimentais demonstram uma redução de 21,8% do repositório original com 99% de precisão e 95% de revocação na identificação de registros duplicados. Além disso, a arquitetura proposta mostrou-se extremamente eficiente e escalável para grandes volumes de dados, deduplicando um repositório de mais de 392 milhões de registros em cerca de uma hora, além de ser de fácil generalização ...

Hardware transactional memory (HTM) systems have been studied extensively along the dimensions of speculative versioning and contention management policies. The relative performance of several designs policies has been discussed at length in prior work within the framework of scalable chipmultiprocessing systems. Yet, the impact of simple structural optimizations like write-buffering has not been investigated and performance deviations due to the presence or absence of these optimizations remains unclear. This lack of insight into the effective use and impact of these interfacial structures between the processor core and the coherent memory hierarchy forms the crux of the problem we study in this paper. Through detailed modeling of various write-buffering configurations we show that they play a major role in determining the overall performance of a practical HTM system. Our study of both eager and lazy conflict resolution mechanisms in a scalable parallel architecture notes a remarkable convergence of the performance of these two diametrically opposite design points when write buffers are introduced and used well to support the common case. Mitigation of redundant actions, fewer invalidations on abort, latency-hiding and prefetch effects contribute towards reducing execution times for transactions. Shorter transaction durations also imply a lower contention probability, thereby amplifying gains even further. The insights, related to the interplay between buffering mechanisms, system policies and workload characteristics, contained in this paper clearly distinguish gains in performance to be had from write-buffering from those that can be ascribed to HTM policy. We believe that this information would facilitate sound design decisions when incorporating HTMs into parallel architectures.

O crescimento na produção e disponibilização de informacões não estruturadas na Web aumenta diariamente. Essa abundância de informações desestruturadas representa um grande desafio para a aquisição de conhecimento que seja processado por seres humanos e também por máquinas. Nesse sentido, ao longo dos anos diversas abordagens têm sido propostas para a extração automática de informações a partir de textos escritos em linguagem natural. Contudo, ainda existem poucos estudos que investigam a extração de informações a partir de textos escritos em português. Diante disso, o objetivo deste trabalhoé propor e avaliar um método não supervisionado para o povoamento de ontologias utilizando a Web como grande fonte de informações, no contexto da língua Portuguesa. Os resultados obtidos com os experimentos realizados foram encorajadores e demonstraram que a abordagem proposta obteve uma taxa de precisão média de 67% na extração de instâncias de classes ontológicas.

Ontologies evolve continuously throughout their lifecycle to respond to different change requirements. Several problems emanate from ontology evolution: capturing change requirements, change representation, change impact analysis and resolution, change validation, change traceability, change propagation to dependant artifacts, versioning, etc. The purpose of this chapter is to gather research and current developments to manage ontology evolution. The authors highlight ontology evolution issues and present a state-of-the-art of ontology evolution approach by describing issues raised and the ontology model considered (ontology representation language), and also the ontology engineering tools supporting ontology evolution and maintenance. Furthermore, they sum up the state-of-the-art review by a comparative study based on general characteristics, evolution functionalities supported, and specificities of the existing ontology evolution approaches. At the end of the chapter, the authors discuss future and emerging trends.

To distribute software, commercial vendors of proprietary software have the opportunity to use some dual licensing (DL) strategy i.e. to provide their software under two different licensing terms (proprietary and open source). We investigate the relevance and impacts of this distribution strategy in the presence of an incumbent open source software competitor. We determine the conditions for this strategy to be profitable for the commercial firm and its impact on price, market shares and welfare. We show that dual licensing may be used as a complement for proprietary software when development spillovers are large. We examine how, in this case, a dual licensing strategy can be used to exclude the open source software from the market and how this is compatible with higher price and lower market share for the proprietary distribution. This situation can also generate conflicts of interests between proprietary software and users resulting in sub-optimal outcomes. Finally, our analysis reveals the key role played by development spillovers and software compatibility for the DL decision.

Resource management is a key factor in the performance and efficient utilization of cloud systems, and many research works have proposed efficient policies to optimize such systems. However, these policies have traditionally managed the resources individually, neglecting the complexity of cloud systems and the interrelation between their elements. To illustrate this situation, we present an approach focused on virtualized Hadoop for a simultaneous and coordinated management of virtual machines and file replicas. Specifically, we propose determining the virtual machine allocation, virtual machine template selection, and file replica placement with the objective of minimizing the power consumption, physical resource waste, and file unavailability. We implemented our solution using the non-dominated sorting genetic algorithm-II, which is a multi-objective optimization algorithm. Our approach obtained important benefits in terms of file unavailability and resource waste, with overall improvements of approximately 400% and 170% compared to three other optimization strategies. The benefits for the power consumption were smaller, with an improvement of approximately 1.9%.

The NSDL Metadata Registry is designed to provide humans and machines with the means to discover, create, access and manage metadata schemes, schemas, application profiles, crosswalks and concept mappings. This paper describes the general goals and architecture of the NSDL Metadata Registry as well as issues encountered during the first year of the project's implementation.

ALMOST HOME IS AN ENTHRALLING EXAMINATION of the pursuit for belonging, the contradictions of black kinship and the contestations of colonial institutions. Ruma Chopra exposes the fraught nature of black mobility and freedom as it sits in uneasy tension with the aspirations of the British colonialists in Jamaica, Canada and Sierra Leone. The picture that emerges is one of the Trelawny Town Maroons as highly strategic actors who negotiated the colonial apparatus of government through conflict, accommodation, and connivance. This is an important contribution to the historiography of the British empire, African-American history, Maroon studies and the emergent work in Black Geographies. Chopra draws on an extensive archive of the Maroons' story both to ensure richness and to demonstrate one of the central themes of the book: Maroons commanded the attention and reflection of the British empire. Maroon existence proved the empire was pregnable rather than a fait accompli. In the introduction we learn that the Maroons signed treaties with the British in 1738 and 1739 that "established the Maroons as useful neighbours" (15). As Chopra shows, though, while the Maroons would at first seem compliant with their role of propping up the plantation economy, they were less ready to accept an inferior racial positioning. War broke out following what Maroons perceived to be an especially demeaning event: the beating of two Trelawny Town Maroons "by a slave overseer the Maroons had previously captured as a runaway; adding insult to injury, the Maroons overheard nearby slaves making jeering remarks" (22). By insisting on these details Chopra highlights the fraught position of the Maroons within a racist colonial system that made their position volatile and their loyalties questionable. The chapter "Bloodhounds" shows the inconsistencies of the perceived social and moral position of the Maroons. As a last resort to ferret out the Maroons from their haunts and retreats, the colonial government brought in Cuban bloodhounds. Yet, as Chopra shows, the coincidence of these devel

Quality of data plays a very important role in any scientific research. In this paper we present some of the challenges that we face in managing and maintaining data quality for a terabyte scale biometrics repository. We have developed a step by step model to capture, ingest, validate, and prepare data for biometrics research. During these processes, there are many hidden errors which can be introduced into the data. Those errors can affect the overall quality of data, and thus can skew the results of biometrics research. We discuss necessary steps we have taken to reduce and eliminate the errors. Steps such as data replication, automated data validation, and logging metadata changes are necessary and crucial to improve the quality and reliability of our data.

Hadoop Distributed File System (HDFS) is the core component of Apache Hadoop project. In HDFS, the computation is carried out in the nodes where relevant data is stored. Hadoop also implemented a parallel computational paradigm named as Map-Reduce. In this paper, we have measured the performance of read and write operations in HDFS by considering small and large files. For performance evaluation, we have used a Hadoop cluster with five nodes. The results indicate that HDFS performs well for the files with the size greater than the default block size and performs poorly for the files with the size less than the default block size.

Product Data Management (PDM) and Software Configuration Management (SCM) are the disciplines of building and controlling the evolution of a complex artifacts; either physical or software. Surprisingly, these two fields have evolved independently; their respective solutions to the same problems are incompatible and their properties are different. PDM is good at modeling while SCM is good at building and supporting concurrent engineering. From a software engineering perspective, the challenge is to take the full potential of strong modeling capabilities, while preserving good concurrent engineering support. The paper shows that rich modeling, flexible evolution, and concurrent engineering supports have conflicting requirements and that a solution requires rethinking the concepts of evolution, versioning and modeling. We have developed a system, called CADSE (Computer Aided Domain Specific Environment), in which a product (software, physical or both) is modeled in a way similar to PDM and in which concurrent engineering and evolution is supported in the SCM way. To that end, the system is driven by models; evolution alone being defined through different models. The paper describes our system and discusses the early lessons of its first years of practical use.

Data warehouse systems integrate data from heterogeneous sources. These sources are autonomous in nature and change independently of a data warehouse. Owing to changes in data sources, the content and the schema of a data warehouse may need to be changed for accurate decision making. Slowly changing dimensions and temporal data warehouses are the available solutions to manage changes in the content of the data warehouse. Multiversion data warehouses are capable of managing changes in the content and the structure simultaneously however, they are relatively complex and not easy to implement. In this paper, we present a logical model of a multiversion data warehouse which is capable of handling schema changes independently of changes in the content. We also introduce a new hybrid table version approach to implement the multiversion data warehouse.

Hardware transactional memory (HTM) systems have been studied extensively along the dimensions of speculative versioning and contention management policies. The relative performance of several designs policies has been discussed at length in prior work within the framework of scalable chipmultiprocessing systems. Yet, the impact of simple structural optimizations like write-buffering has not been investigated and performance deviations due to the presence or absence of these optimizations remains unclear. This lack of insight into the effective use and impact of these interfacial structures between the processor core and the coherent memory hierarchy forms the crux of the problem we study in this paper. Through detailed modeling of various write-buffering configurations we show that they play a major role in determining the overall performance of a practical HTM system. Our study of both eager and lazy conflict resolution mechanisms in a scalable parallel architecture notes a remarkable convergence of the performance of these two diametrically opposite design points when write buffers are introduced and used well to support the common case. Mitigation of redundant actions, fewer invalidations on abort, latency-hiding and prefetch effects contribute towards reducing execution times for transactions. Shorter transaction durations also imply a lower contention probability, thereby amplifying gains even further. The insights, related to the interplay between buffering mechanisms, system policies and workload characteristics, contained in this paper clearly distinguish gains in performance to be had from write-buffering from those that can be ascribed to HTM policy. We believe that this information would facilitate sound design decisions when incorporating HTMs into parallel architectures.

In WebDAV: Next-Generation Collaborative Web Authoring, Lisa Dusseault thoroughly describes the WebDAV protocol and the rationale behind the current version (see Y. Goland et al., HTTP Extensions for Distributed Authoring WebDAV (http://asg.web.cmu.edu/rfc/rfc2518.html), RFC 2518, Feb. 1999). Dusseault explains the WebDAV protocol's strengths and weaknesses with remarkable clarity. WebDAV stands for Web-based distributed authoring and versioning, although the standard doesn't include versioning support itself. The DeltaV standard, which is defined in a different RFC document (G. Clemm et al., Versioning Extensions to WebDAV (http://www.ietf.org/rfc/rfc3253.txt), RFC 3253, Mar. 2002), addresses versioning issues separately.

Deduplicação de registros (DR) tem como objetivo identificar instâncias que representam a mesma entidade do mundo real em repositórios de dados. No ambiente governamental, o processo de DR facilita a identificação de irregularidades e reduz o consumo de recursos computacionais em tarefas de integração de dados. Nesse contexto, propomos neste artigo uma plataforma escalável, efetivo e eficiente para integração de grandes repositórios de dados (i.e., com grandes volumes de dados, da ordem de milhões de registros) para unificar entidades duplicadas oriundas de múltiplas e diferentes fontes. Nossos resultados experimentais demonstram uma redução de 21,8% do repositório original com 99% de precisão e 95% de revocação na identificação de registros duplicados. Além disso, a arquitetura proposta mostrou-se extremamente eficiente e escalável para grandes volumes de dados, deduplicando um repositório de mais de 392 milhões de registros em cerca de uma hora, além de ser de fácil generalização ...

This paper addresses the design and implementation of an adaptive document version management scheme. Existing schemes typically assume: (i) a priori expectations for how versions will be manipulated and (ii) fixed priorities between storage space usage and average access time. They are not appropriate for all possible applications. We introduce the concept of document version pertinence levels in order to select the best scheme for given requirements (e.g., access patterns, trade-offs between access time and storage space). Pertinence levels can be considered as heuristics to dynamically select the appropriate scheme to improve the effectiveness of version management. We present a testbed for evaluating XML version management schemes.

In the rapidly evolving Cloud market, the amount of data being generated is growing continuously and as a consequence storage as a service plays an increasingly important role. In this paper, we describe and compare two new approaches, deriving from the EU funded FP7 projects OPTIMIS and VISION Cloud respectively, to filling existing gaps in Cloud storage offerings. We portray the key value-add characteristics of their designs that improve the state of the art for Cloud computing towards providing more advanced features for Cloud-based storage services.

Active Storage provides an opportunity for reducing the bandwidth requirements between the storage and compute elements of current supercomputing systems, and leveraging the processing power of the storage nodes used by some modern file systems. To achieve both objectives, Active Storage allows certain processing tasks to be performed directly on the storage nodes, near the data they manage. However, Active Storage must also support key requirements of scientific applications. In particular, Active Storage must be able to support striped files and files with complex formats (e.g., netCDF). In this paper, we describe how these important requirements can be addressed. The experimental results on a Lustre file system not only show that our proposal can reduce the network traffic to near zero and scale the performance with the number of storage nodes, but also that it provides an efficient treatment of striped files and can manage files with complex data structures.

The capability of taking snapshots is approaching ubiquity as a feature of file systems and data storage arrays. Here, we present an approach to structuring and managing snapshots in a storage space that provides for rapid creation and roll-back. This approach has been realized in the form of a Plan 9 kernel device that can be interposed between any pair of storage service and application. The kernel device has, in turn, been used to support an experimental file system. In this paper, we discuss the basic snapshot model, its implementation, and its application to a file system. Finally, we consider extensions to the model for supporting the deletion of snapshots.

O sistema judiciário é composto por inúmeros documentos relacionados a processos jurídicos. Esses documentos podem conter informações relevantes que suportem a tomada de decisão em processos futuros. No entanto, a coleta dessas informações não é uma tarefa trivial. Este artigo propõe o uso de agrupamento para reunir processos semelhantes e facilitar a coleta de informações. Dessa forma, diferentes abordagens foram avaliadas com a intenção de identificar a mais adequada para realizar esta tarefa. As abordagens foram aplicadas a uma base de dados composta por 1515 textos de fatos de petições iniciais. Essas abordagens foram avaliadas levando em consideração métricas de avaliação internas e os textos dos processos agrupados. Os resultados apontaram que a melhor abordagem para realizar o agrupamento de processos jurídicos é composta pelo algoritmo K-Means e pela técnica de representação TF-IDF em combinação com a técnica PCA.

In this paper we discuss several features of XP we have used in developing curricula and courses at Duke University and the University of Northern Iowa. We also discuss those practices of XP that we teach as part of the design and implementation process we want students to practice as they develop programming expertise and experience. In theory the academic study of programming and software development should be able to embrace all of XP. In practice, however, we find the demands of students and professors to be different from professional and industrial software developers so that while we embrace the philosophy and change of XP we have not (yet) adopted its principles completely.

We present NeST, a flexible software-only storage appliance designed to meet the storage needs of the Grid. NeST has three key features that make it well-suited for deployment in a Grid environment. First, NeST provides a generic data transfer architecture that supports multiple data transfer protocols (including GridFTP and NFS), and allows for the easy addition of new protocols. Second, NeST is dynamic, adapting itself on-the-fly so that it runs effectively on a wide range of hardware and software platforms. Third, NeST is Grid-aware, implying that features that are necessary for integration into the Grid, such as storage space guarantees, mechanisms for resource and data discovery, user authentication, and quality of service, are a part of the NeST infrastructure.