专业英语论文-云计算和数字图书馆.doc

- 9 -Cloud Computing Technology and Digital LibraryAbstract:As a rapidly developing new information technologies,Cloud computing has been applied in many fields at home and abroad，it will also help the construction and development of the Digital Library bringing the impact which can not be ignored. On the basis of introducing the principle of cloud computing,discuss the influence of Cloud computing on the University Digital Library and some problems which should be pay attention to 。 The article firstly analyzes the concept of cloud computing， the concept and difficulties of digital library， and the cloud on the Digital Library from the information resources， information users, information workers and information facilities. In the end， it prospects the development for the future. Cloud computing has brought us a new perspective to look at the current resourcesharing problem， cloud computing can be applied to digital library resources to improve information sharing capabilities and improve resource utilization. Our results show that a cloud implementation may be a feasible alternative for its continued operation and growth.Key Words: Cloud Computing, Digital Library， Cloud Computing ApplicationsIntroduction section With the popularity of Internet and upgrading network technology , It has quietly entered our lives becoming a part of peoples lives ， changing our way of life 。 We used learn and acquire knowledge in the classroom， books and libraries 。 Now we can query the information we want at comfortable home 。 Computer technology brought fresh blood to the old library science ，also brought inevitable problems. Today， the shortcomings malicious attack or a single server mode will lead irreparable damage to the lost of libraries important information . Now we will leverage cloud computing technology to help avoid loss of digital libraries , and provide good service for readers better and more humane way 。 Digital Library is a thing produced certainly with the development of the Internet,the advantage is obvious compared with the traditional library。 It makes the Information storage space greatly reduced without damage, and Information retrieval more convenient。achieving the target of remote transmission of information and information Sharing，so all colleges and universities are actively building the digital library 。 However， Current digital Library is faced with a series of problems such as security risks， difficulties in system maintenance and large capital investment，leading to the development of digital Library experienced difficulties。Technology is advancing rapidly，nowadays， the emergence of cloud computing technology has brought new hope and vitality to the Digital Library, it will help digital library out of dilemma，and provide the necessary services for readers better 。 Part 1 Cloud Computing OverviewCloud computing is a distributed processing, parallel processing and grid computing development，it aimed to integrate numbers of lowercost computing entities into one system with a supercomputing power through the network。 In the remote data center,tens of thousands of computers and servers connected into a computer goes to work，making the related calculation distributed in the large numbers of distributed computer rather than the local computer。 Therefore， The purpose of the cloud computing is depend on B / S Architecture, transferring the calculation of the pressure from the client to the server。 Users only need to pay little service fee can be configured computers， laptops， cell phones， etc accessing to the ”cloudcenter， operationing according to their own needs.The implementation of cloud computing will bring us more computing power, lower costs and peopleoriented service。 Cloud computing as a relatively new concept， once appranced can be pursuit of industry and academia。 In particular， since 2007， including Microsoft, IBM， GOOGLE, etc。, the major company has successfully launched a cloud computing service,which can provide users with online data storage， information transfer， large amounts of data parallel processing， data indexing and query services ， and brought great benefits 。 In China， the development of cloud computing is very rapidly， a number of Cloud Computing Center has established ， developed a variety of cloud computing products, servering in various fields and levels of users . National library community are closely watching the cloud， Library Society of China and the Shanghai Society for Library Science Academic Committee held symposium several times about this related topics of ”cloud computing and library” 。 1。1 The characteristics of cloud computing1， Large scale 。 ”Cloud” has a considerable size . Google cloud computing already has more than 100 million servers 。 Amazon， IBM， Microsoft， Yahoo and others ”cloud has hundreds of thousands of servers 。 Private companies typically have hundreds of thousands of cloud server。 Cloud” can give users an unprecedented computing power 。 2, Virtualization 。 Cloud computing allows users in any location, using a variety of terminal accessing to application services 。 All the resources requested are from the cloud” , instead of fixed physical entities 。 Application is running in the somewhere of ”cloud” 。 but in fact users do not understand and worry about the specific location of running applications 。 Only need a laptop or a mobile phone , web services can achieve all we need ， even the task of Supercomputing . 3， High reliability 。 ”Cloud” uses the measures such as multiple copies of data fault tolerance， computing nodes are interchangeable with the structure to protect the reliability of services . making cloud computing more reliable than using the local computer 。 4, Universal 。 Cloud computing is not the specific application , ever-changing application can be constructed under the support of the of ”cloud ， with a cloud” can support different applications running 。5， High scalability 。 Cloud”s size is scalable dynamically ， meeting the needs of the scale of growth of Applications and users 。 6， Ondemand service 。 Cloud” is a huge resource pool ， You should buy on demand 。 Cloud can be billed like running water, electricity and gas .7, Extremely cheap . Very low-cost nodes can be used to form the clouds as the special fault tolerance measures of cloud” 。 Cloud”s centralized and automated management make large numbers of business without the burden of the high cost of data center management 。 ”Cloud of the universal resource utilization improve significantly compared to traditional systems 。 So users can fully enjoy the low-cost advantage of ”cloud” , often only takes a few hundred dollars or a few days to complete the task used to take tens of thousands of dollars, several months to complete 。 Part2 Background and related workCloud computing and infrastructure have been discussed in the context of information retrieval systems which are related to grid computing and distributed computing 。 In these discussions， the focus has been either on cloud computing for data storage infrastructure or the compute infrastructure。 A discussion of cloud computing for scalability in preprocessing， harvesting， transformation and storage is provided by in the context of crawling the Web with Sindice. In the following sub sections, we briefly overview the architecture of the SeerSuite application framework and its deployment as CiteSeerx. This section is meant to provide an understanding of the application requirements that are crucial for understanding the challenges and the need for infrastructure virtualization and abstraction。Figure 1 provides an overview of SeerSuite architecture。SeerSuite includes components of both web search systems and digital libraries 。 SeerSuite consists of several components and services loosely coupled together with REST and SOA interfaces. We group components of the framework into Web application, data storage, extraction，ingestion, and maintenance systems。 These components can function as standalone applications.Fig。 1. SeerSuite Architecture1） Web Application： The primary interface between users and SeerSuite is the Web application。 Users can search，browse, and traverse the collection through this interface。Queries in SeerSuite are supported through an interface over REST to an index, an inverted file data structure which allows for fast and accurate retrieval。 Results obtained from the index are processed by the application and with the database displays search results. Users can view document metadata from summary pages. These summary pages allow a user to download a cached copy or be directed to another source, view citations from the document， view tags created by users， andtrack changes to the document。 The autonomous citation indexing feature enables users to rank documents based on citations to the document inside the collection as well as browse the collection through citation links。 These features require the citation graph structure of the collection to be stored in the database。 In addition to providinguser access through web pages, SeerSuite supports an Open Archives and an Application Programming Interface， which provide programmatic access to the database。 CiteSeerx supports myCiteSeer, a personalization portal which with support of the database, allows users to store queries， document portfolios, tag documents and make corrections to document metadata.2） Crawling and Metadata Extraction: The document acquisition process of SeerSuite includes two major tasks。 The first is to obtain documents relevant to the collection。 A focused crawler traverses the web to find documents relevant to a particular topic。 In SeerSuite， a focused crawler scans web pages from a crawl list and fetches documents， mostly PDF， embedded in these pages. Before metadata extraction can take place, the incoming documents are filtered and converted into text by filtering and document conversion modules。 Conversion modules use standalone off the shelf applications such as PDFTet or PDFBox for converting documents。 SeerSuite adopts state of the art metadata extractionMethods to extract document metadata using heuristics and machine learning methods。 The header parser extracts document title， author, abstract and affiliation information。 A reference parser， ParsCit extracts reference information including citations and citation metadata. Documents can be processed individually or through a batch process， using Business Process Execution Language（BPEL）, orchestration of individual service oriented modules。3) Ingestion： Processed documents from the extraction system are ingested into the collection. In the first step, the metadata extracted from the documents and metadata from the crawl are aggregated into a single file, which serves as the source for updates to the database. Ingestion involves the database， the repository and the extraction systems. This step adds information about the incoming document and embedded objects to the database and updates the citation graph structure。 The document metadata, processed files with metadata extracted by different modules， and original document are placed in the repository.4） Maintenance： Updates to the index， generating charts and statistics are part of the maintenance service. The maintenance flow generally involves the database, the index and the repository。5） Federated Services： While SeerSuite offers several features for document/citation search and indexing, services such as table, algorithm， figure search are also valuable to users. However， such services may have their own stack of components and interfaces. The table search feature in CiteSeerx is one such service。 The components are hosted along with existing CiteSeerx services, but the interfa e and storage services for table search are independent of the main index and database. Such a federated approach is useful for services still being researched or developed。6） Backup and Replication： CiteSeerx maintains both offline and onsite backups for the repository, database and index。Part3 The benefits of cloud computing bringing for the university digital library Cloud computing is hot in todays network technology ， its emerging pattern of shared infrastructure ， numerous systems pool can be connected together providing a variety of digital services . After digital library in the experience of the Internet era, WEB times, the grid era， WEB2。0 ， it slowly into the cloud computing era 。 Cloud computing technology applicated to the digital library will bring a real transition phase , and profound impact on library services 。 We believe cloud computing will bring changes for the library industry 。 In recent years， colleges and universities are working on digital libraries direction,most of them also built a primaryscale digital library，but it is premature to think it has formed an overall scale, efficiency。As the construction of the digital library currently exists a range of outstanding issues. Today， the application of cloud computing technology can help digital library solve problems and overcome difficulties 。 3。1 Cloud”to ensure the safety of library。Digital library generally has a huge information resources， Digital libraries are the current focus on the data stored in a server set up , once the server has a fault or malicious attack, it will make user information and library data loss， thereby causing irreparable consequences for digital library .Cloud computing technology can ensure the reliability of digital library resources stored，avoiding errors caused by the server resources Loss。 At the same time,it ensure in unsatisfactly network security environment in the event of loss of information,digital libraries can continue to provide services for users . This is because there are numerous servers in the ”cloud , Even if there is a server problem，Then the other servers can transfer backed up data quickly in a very short period of time to another server，and start the server to provide services ， avoiding problem which a server in the network paralysis after prolonged exposure to attack , so that digital library achieve uninterrupted security services in the real way. 3.2 Reduce the cost of building the digital library In todays society the development of electronic devices technology is rapidly， to meet this trend ， making the Library continue to operate normally , needing to continue to invest funds for the upgrading of hardware， maintenance of database systems and hardware also requires large cost。 In the environment of cloud computing ， general library users simply use the digital library system as the end user , And this configuration of the terminal device does not require high , there is no need to install their own softwa