Z39.50

from Wikipedia, the free encyclopedia
Z39.50
Family: Internet protocol family
Operation area: Query information systems,
portals,
bibliographies, etc.
Port: mostly 210 / TCP
Z39.50 in the TCP / IP protocol stack :
application Z39.50
transport TCP
Internet IP ( IPv4 , IPv6 )
Network access Ethernet Token
bus 		Token
ring 		FDDI ...
Standards:

ISO 23950
ANSI / NISO Z39.50-2003

Z39.50 is a network protocol that is used in libraries as the standard for querying bibliographic information systems. The protocol is a connection-oriented protocol on application layer 7 of the OSI reference model . Z39.50 generates one or more sessions ( simultaneous sessions ) between Z39.50 clients and servers ( up to Z39.50 Version 3 in OSI terminology: Origin and Target ). A transformation syntax enables the uniform query of heterogeneous information systems . The respective information system can transmit the data in different forms so that it can meet specific requirements.

Z39.50 is predestined for use in portal environments that enable parallel searches in several bibliographic information systems via a uniform access.

However, Z39.50 is also widely used in reference management programs such as EndNote , LibraryThing or learning management systems such as Stud.IP , which use the contents of bibliographic information systems to create your own literature lists.

Administration and development of the protocol is carried out by the Library of Congress , which has been the Maintenance Agency Z39.50 since 1990 . The successor to Z39.50 is the HTTP-based protocol Search / Retrieve via URL (SRU / SRW), which was developed as part of the Z39.50 International Next Generation (ZING) project.

history

development

The development of Z39.50 was started in 1984 in the USA by the National Information Standards Organization (NISO) and has been continued by the Library of Congress since 1990. Several versions have been implemented:

  • Z39.50-1988, Z39.50-1992 ( version 2 ), Z39.50-1995 ( version 3 )
  • Z39.50-2003 ( version 4 ) as the current version, which includes versions 2 and 3 and essentially represents a consolidation of the version 3 standard.

In addition to the US standards ( ANSI / NISO Z39.50-2003 ), Z39.50 has also been International Standard ISO 23950 since 1998: "Information Retrieval (Z39.50): Application Service Definition and Protocol Specification".

The large-scale introduction of Z39.50 in Europe took place within the scope of the ONE - Opac Network Europe project until the end of the 1990s, in Germany through the DBV-OSI project (1993–1997) and is almost exclusively limited to bibliographic information systems.

DBV-OSI

In Germany, the DBV-OSI ( German Library Association - Open Systems Interconnection ) project was carried out from 1993–1997 with the aim of expanding the bibliographic information systems with a Z39.50 interface , funded by the DFG and the federal government . After a preliminary project, the realization took place in stage II (DBV-OSI-II).

Project partners were:

The following Z39.50 basic services were implemented: INITIALIZE , SEARCH , PRESENT , DELETE_RESULT_SET , RESOURCE_REPORT , SCAN , CLOSE .

The result of DBV-OSI is the widespread use of Z39.50 for querying data from bibliographic information systems in Germany (status 2007).

Protocol specification

The detailed protocol specification can be found on the Library of Congress website:

The essential stipulations of the Z39.50 protocol are:

  • Protocol services
  • Transfer syntax ( operators, attribute sets )
  • formats used ( record syntaxes )

Log services

Brief description of the basic services of the DBV-OSI project:

INITIALIZE - authentication of the client, opening of a Z39.50 session

  • Protocol-specific specifications for the session: including protocol version, applicable services, preferred and maximum message size

SEARCH - search queries

  • Client: Query, name of result sets and format
  • Server: number of hits, status information, possibly some data records

PRESENT - transmission of search results

  • Client: Name of result set and format, numbers of the records
  • Server: transfers desired records

DELETE_RESULT_SET - Deletion of result sets in the target system

  • Client: Name of the result set (s)
  • Server: Delete operation status

RESOURCE_REPORT - Exchange of accounting information

  • Client: Query of the costs incurred (resources used)
  • Server: costs incurred

SCAN (BROWSE) - Search in ordered term lists (e.g. title, author, and so on)

  • Client: Term list, start term, size of the scan steps
  • Server: scan status, number of hits

CLOSE - closing the Z39.50 session (from V3)

  • can be initiated by client or server
  • Server: Deletion of all result sets created during the session

Transfer syntax

Because of the different query syntaxes for client and server, the query languages ​​are transformed into the system-specific syntax on both sides. Various query types are defined for this in the Z39.50 protocol specification ( RPN , CCL, etc.).

Example:

  • Query client: SEARCH per Beckmann, Andreas AND tit Gott Bautzen
  • Query server: FIND AU = Beckmann A AND TIT = Gott Bautzen

By transferring the search terms into a number system ( attribute set ), the problems of upper / lower case ( tit → TIT ) and the different names with the same semantics ( per → AU ) are solved.

The above query is then formulated as follows using the attribute set (Bibliographic Information) in the query language PQF (Prefix Query Format):

  • find @and @attr 1 = 1 @attr 5 = 100 "Beckmann, Andreas" @attr 1 = 4 @attr 5 = 100 "Gott Bautzen"

Formats used

The servers offer the clients the data in at least one, but generally in several formats. Typical bibliographic formats ( record syntaxes ) for Z39.50 servers are:

implementation

Target Profiles

The use of the attributes for the target systems (Z39.50 server) is described in so-called target profiles. The following information is usually contained in a target profile:

  • Type of access (public or, for example, with IP restrictions, possibly user and password information)
  • Name of the database
  • used fonts
  • used ports (often several ports for different character sets)
  • Name of the host for access (domain name)
  • List of searchable fields with assigned attribute set

Access via a web browser

There are basically two options for accessing a Z39.50 server via a web browser:

1) direct access via browser-integrated plug-in

There are approaches here, for example for Netscape (add-on) and MS Internet Explorer (Active-X). However, these should hardly have been used for any application.

Fig .: HTTP-Z39.50 gateway

2) Access via HTTP Z39.50 gateway

The standard solution for accessing a Z39.50 server is via HTTP via an HTTP-Z39.50 gateway, which receives the HTTP requests from the client, extracts the Z39.50 protocol from them and sends them to the selected Z39.50 server forwards for processing. Conversely, the Z39-50 responses from the server are converted back into HTTP responses and transmitted to the client.

Implemented and freely accessible HTTP Z39.50 gateways are for example:

Software and toolkits

The Z39.50 Maintenance Agency of the Library of Congress provides an extensive list of software and toolkits on its website that support the integration of the Z39.50 functionality into your own clients or server systems.

perspective

From the perspective of 2007, 10 years after the end of the DBV-OSI project, various developments have occurred that can take over the functionality of Z39.50 in whole or in part.

These developments are characterized by the following features, among others:

  • Functionality of Z39.50 is brought into the WWW ( HTTP )
  • Use of sufficiently powerful query languages ​​( CQL )
  • more variable to structure and easier to process ( XML , XSL etc.)
  • formats: Development of corresponding XML counterparts ( MARCXML , MABXML etc.)
  • easier to implement (without complex transfer syntax)
  • Supported by the formation of large library networks , global standards for regulations and formats, sufficient bandwidths in wide area networks , as well as the provision of open interfaces for querying metadata ( e.g. OAI-PMH ), the merging of databases is made possible - the importance of querying distributed information systems with Z39.50 will decrease with it.

This includes the following developments / projects:

ZING - Z39.50 International Next Generation

Examples :

Metadata Harvesting (OAI-PMH)

  • Harvesting ( collecting ) the metadata (protocol: OAI-PMH ) from various information systems ( data providers ) makes it possible to save this data in your own information system, process it and then make it available in a normalized, enriched form tailored to the application . The Open Archive Initiative - Protocol for Metadata Harvesting protocol used is HTTP-based and delivers the data in XML. The standard data format is Dublin Core .
  • With the merging of the data, querying the distributed information systems becomes superfluous
  • the OAI-PMH is already being used for data exchange between network systems
  • See also the recommendations of the Science Council on the provision of digital information by university libraries (Drs. 4935/01) and DINI certificate for document servers

Example :

Link resolver

  • The HTTP-based service link resolver enables the consolidation of distributed resources in an "intelligent" service portal. For the search for publications using metadata of the source (source) as well as of stored information about local stocks and licenses of the target (target), links to "suitable" services such as electronic full texts, proof in catalogs, document delivery service, WEB services such as ISI Journal Citation Report, WEB of Science and others generated.

Example :

Merging of bibliographic data stocks

  • The concentration of the libraries in associations and the worldwide establishment of standardized formats and sets of rules for the construction of bibliographic information systems enable the physical and / or logical consolidation of previously distributed heterogeneous data stocks in a few central, largely homogeneous information stocks.
  • Union catalogs, national catalogs and a world catalog are created.
  • As the data is merged, the importance of querying distributed information systems with Z39.50 will decrease.

Example :

  • Cooperation initiative hbz - BVB - OBVSG. Investigation to merge the three central network databases. Final report (PDF; 153 kB). ( discontinued in June 2007 )
  • "Cooperative new cataloging" project
Barbara Block, Josef Labner, Beate Rusch: Cataloging made cooperative. ( Memento from May 12, 2012 in the Internet Archive ) (PDF; 162 kB)
In: Library Service, Volume 41 (2007), no. 2
z. B. Zubarev, Morozov, Röpke. "Statistical Mechanics of Nonequilibrium Processes: Basic concepts. Kinetic theory."

Web links

Individual evidence

  1. ↑ Bib-1 Attribute Set ( English ) loc.gov. Retrieved September 14, 2019.