E-LIB Bremen

from Wikipedia, the free encyclopedia

The E-LIB ( Electronic Library of the State and University Library Bremen ) is a virtual library that has existed since 1998 , which makes both the physical library inventory of the SuUB Bremen and free or licensed electronic media available under a search interface. In addition, the E-LIB provides a document server that authors from all Bremen universities can use to publish scientific papers as Open Access . The main target group of the E-LIB are the members of the University of Bremen and the universities in the Free Hanseatic City of Bremen .

history

In 1996 the DFG project OSIRIS (Osnabrück Intelligent Research Information System) began. This project was carried out jointly by the University of Osnabrück and the associated university library (UB). A database was developed that combined access to specialist information and specialist databases under one search entry. This was done initially with a view to the individual needs of the Osnabrück University Library, but with the intention that other users could take over this application. The OSIRIS and E-LIB projects started in Osnabrück were used and further developed in Bremen from 1998/1999. Initially, the E-LIB was designed as a pure verification system for electronic media. From 1999 on, this project experimented with search engines and harvesting. Osiris served as a research method for the printed titles of the SuUB. The focus here was on the thematic entry points. In 2004 both projects were merged. From this point in time, users had a uniform search interface for all media types, while in other libraries the separation of printed and electronic media was standard and in some cases continues to this day (2009). In February 2006 the E-LIB document server was certified by the German Initiative for Network Information (DINI) . It is one of the 33 document servers that have acquired this certificate to date (as of November 17, 2009). In the same year, the E-LIB was awarded the “Innovation Prize for trend-setting library IT ” and a module for integrating metadata from Open Access offers was introduced. The technical basis for this is developed at the SuUB itself. It is possible to access three million full-text documents, the scope of which is continuously expanding. From the E-LIB, the user can access nationally licensed journals, for example from the publishers Elsevier and Springer . With a relaunch in 2008, a. technical improvements of the search functionality realized, such as B. a search interface with drill-down techniques. By integrating external Web 2.0 services such as Wikipedia , LibraryThing and Google Books , new usage options are continuously being opened up for users of the E-LIB Bremen.

Content

The aim of the system is to make available all stocks that are locally and externally available to Bremen users in a single retrieval system. In January 2011, the media offer was put at 32 million, including over 38,000 e-books .

In addition to the media listed in the library catalog of the Bremen State and University Library, the following media can also be searched using the E-LIB search engine:

  • all printed and electronic journal titles that the library owns or for which it provides access (approx. 31,000)
  • a collection of currently around 38,000 e-books
  • the start pages of all bibliographic specialist databases to which the library provides access (approx. 200)
  • Electronic magazine articles from major international publishers (approx. 23 million)
  • scientifically relevant free internet sources and freely accessible documents on university document servers worldwide (approx. 3 million documents)

The high percentage of electronic full texts, which make up around 80% of the content, is striking.

technical basics

The technical basis for the E-LIB is the CiXBase System software, which is a further development of the Osnabrück Intelligent Research Information System (OSIRIS).

With their help, the heterogeneous metadata that are distributed everywhere are first imported into the E-LIB system ( data warehouse solution). To build a consistent data pool, they are converted into a homogeneous XML flat file structure in the next step . This describes a file system without subdirectories; all data are here on one level directly behind one another.

The now converted XML data are then indexed using search engine technology. So-called statistically balanced B-tree algorithms are used. The search is controlled via nodes so that when a query is made, the entire index is not searched from front to back, but the destination is reached via five to six stations.

This local index (“local indexing”) is built up overnight in around eight hours a day, so that the system is always kept up to date with a one-day delay. The main advantages of such a local index are

  • fast response times
  • a uniform representation
  • a balanced hit display (ranking)
  • the possibility of using the metadata of the search results for further subject-specific analyzes that can be carried out during the search

The disadvantage is that some data is lost due to the normalization.

Full text indexing has been deliberately avoided in order to avoid precision problems, since only a fraction of the full texts are available on the company's own server and are therefore not available. In addition to import, conversion and indexing, other services are offered via the CiXbase system:

  • balanced relevance ranking
  • Connection of LDAP servers for the exchange of user data
  • Support of SRU (Search / Retrieval via URL), for example for the "no hit analysis"
  • Spell check (“Did you mean”) for typos based on the Levenshtein distance algorithm
  • Formation of "tag clouds" based on computer linguistic and statistical analyzes

The presentation of the data was developed by the SuUB itself. The XML-structured metadata is displayed in the user's browser application using XSLT stylesheets. This display has also been expanded to include other services, for example mashups (e.g. via SeeAlso API: Google Book Search, LibraryThing, Wikipedia).

The separation of data structure and layout means that the metadata is almost completely interoperable with other systems. The modular technology of the search engine and services makes it possible to continue to use parts of the E-LIB in a different context (e.g. generation of tag clouds).

Other techniques used are:

  • OAI harvesting technologies with the help of which metadata, which are provided on document servers, are collected and integrated into the E-LIB.
  • Discovery search: an Ajax -based recommendation function ("our customers' favorites")

Comparable systems

Since the E-LIB goes beyond traditional OPACs in terms of technology, scope and usability, it is also counted among the “Next Generation Catalogs” (also “Catalog 2.0”) in specialist library circles. In Germany, Austria and Switzerland there are some comparable systems, some of which are available as open source. These include the in-house developments XOPAC ( University Library Karlsruhe ), OpenBib ( University Library Cologne ), HEIDI ( University Library Heidelberg ), BASE ( University Library Bielefeld ) and the Virtual Institute Library Information Science ( University of Saarland ), various projects based on Solr / Lucene and / or Vufind such as Beluga ( libraries in Hamburg ) and the Suchkiste ( joint library network ), as well as the commercial products Primo and Touchpoint .

literature

Web links

  • E-LIB Electronic Library State and University Library Bremen

Individual evidence

  1. ↑ Marc Ronthaler, Hartmut Zillmann: literature review with OSIRIS. A test of the OSIRIS retrieval component . In: Library Service . tape 32 , no. 7 , 1998, pp. 1203-1212 .
  2. ↑ DINI-certified servers. Archived from the original on April 7, 2012 ; Retrieved February 3, 2011 .
  3. ↑ “eUniversity” awards innovation prize for library IT to Hamburg, Bremen and Bielefeld . In: Library Service . tape 40 , no. 12 , 2006, p. 1461-1461 .
  4. ^ A b Martin Blenkle, Rachel Ellis, Elmar Haake: E-LIB Bremen - Automatic recommendation services for specialist databases in the library catalog / metadata pools as a knowledge base for services that are independent of the inventory . In: Library Service . tape 43 , no. 6 , 2009, ISSN  0006-1972 , p. 619 .
  5. ↑ Project information E-LIB. Retrieved November 16, 2011 .
  6. ↑ E-LIB Search Engine - Frequently Asked Questions (FAQ). Retrieved February 3, 2009 .
  7. ↑ CiXBase system. (No longer available online.) Formerly in the original ; Retrieved December 11, 2009 .  ( Page no longer available , search in web archives )  Info: The link was automatically marked as defective. Please check the link according to the instructions and then remove this notice.@1@ 2Template: Dead Link / cixbase.dyndns.org  
  8. ↑ ( Page no longer available , search in web archives: Osnabrück Intelligent Research Information System (OSIRIS) )@1@ 2Template: Dead Link / deposit.ddb.de
  9. ↑ Next Generation Catalogs in Europe. Retrieved December 15, 2009 .