DIE ÖSTERREICHISCHE BIBLIOTHEKENVERBUND UND SERVICE GMBH
DIE KONSORTIALE PRIMO-IMPLEMENTIERUNG IM OBV
Teil 1. Aktueller Stand, Herausforderungen / Lösungen (V. Babitchev, J. Brandauer) Teil 2. Integration zusätzlicher Dienste (U. Krabo)
DACHELA 2010, Wien, 15. Juni 2010
Inhalt
Der aktuelle Stand
- Was haben wir geschafft?
- Datenmodell - Datenfluss - Aktuelle Kennzahlen
- Dienstleistung
- Die neuen Teilnehmer
Herausforderungen / Lösungen Generelle Herausforderungen Online-Demo / Lösungen
- Alle Daten für eigene Sicht - Real Time Availability – RTA
- Mehr Titel als im eigenen Lokalsystem – IVSCAN Versorgung - Abstracts in Vollanzeige – „Plug In 1“
- Volltext-Implementierung – „Plug In 2“
Produktionsablauf – Zusammenspiel der Komponenten
DACHELA 2010, Wien 2
Was haben wir geschafft?
Aufbau der technisch / organisatori- schen Infrastruktur
3 Primo-Sichten in Produktion Volle Integration der eDOC-Daten Integration von MetaLib / SFX
Anreicherung der IVSCAN-Aufsätze Authentifizierung über Shibboleth Web-2.0-Funktionen + Gadgets
DACHELA 2010, Wien 3
Datenmodell / Datenfluss
DACHELA 2010, Wien 4
HOL SE BIB01
Z30
ADM
BIB HOL Z300
Publish zentral Aufberei- tung PPS
eDOC
Publish Lokal 1
SE
BIB02 Publish Lokal 2
UBW ACC UBI
Lokalsystem Verbundsystem PRIMO
ML KB SFX Digit Repos
Normalisie- rung
Enrich- ment
Laden
Dedup FRBR
Plug-in-1
Harvesting
Lokale BIB- Felder nach HOL
Norm
Plug-in-2
Exporte für PRIMO PNX
Aktuelle Kennzahlen Stand: 17. Mai 2010
IVSCAN: Titel die aus dem Verbund in den lokalen Primo-Bestand transportiert werden (ca. 650.000 Aufsatztitel)
eDOC: Anreicherungsdaten aus eDOC (ca. 420.000 Objekte) Lokale BIB: Lokale Daten (keine AC-Nummer) aus UBI01
Verbund BIB IVSCAN eDOC Lokale BIB SFX MetaLib Summe
UBI 1.481.354 272.411 132.985 226.219 42.254 75 2.022.373
UBW 3.000.875 455.131 183.606 62.020 401 3.518.427
ACC 7.619.183 - 301.965 - - 7.619.183
Summe 12.101.412
DACHELA 2010, Wien 5
Dienstleistung
Realisierung eines standardisierten Workflows zur Primo-Einführung
Ziel: Minimierung der Aufwände neuer Teilnehmer
- Basiseinrichtung Primo - Einrichtung PPS
- Datenvorbereitung Aleph
- Pflege des Aleph-Standard-Templates - Projektmanagement
- Front-End-Schulung
Neuer Teilnehmer konzentriert sich auf - Test der Normalisierung / Funktionalität - Web-Gestaltung
- Integration und Bereitstellung lokaler Datenquellen
Alle Entwicklungen können von allen Teilnehmern genutzt werden!
DACHELA 2010, Wien 6
Die neuen Teilnehmer
Einrichtung Produktion Version
Veterinärmedizinische Universität Wien 2010 Q. 2 2
Migration auf 3
Wirtschaftsuniversität Wien 2010 Q. 3 - 4 3
Universität Graz 2010 Q. 4 3
Universität für angewandte Kunst 2011 Q. 1 3
Österreichische Nationalbibliothek 2011 Q. 1 3
DACHELA 2010, Wien 7
Issues in implementing Primo using Central data
8
„Central Catalog is the primary data source for local Primo views“
Central bibliographic data – OK!
– common bibliographic fields – authority files
The challenge is local bibliographic fields that are stored in local systems!
• Local bibliographic fields (our solution)
– cataloged now in HOL and not in BIB records (local subjects etc.) – transported to the Central Catalog by Aleph data replication
– integrated into Central bibl. record by Aleph Publishing Mechanism
Issues in implementing Primo using Central data (cont.)
9
Problems to be solved
• No efficient solution for producing Primo data sources from the Central Catalog for institutions
− “standard” solutions assume big data redundancies
• Central records do not store local system numbers
− the mandatory element in implementing the real time availability (RTA) _ for institutions views
− the base for implementing back links to local OPACs
Issues in implementing Primo using Central data (cont.)
10
Challenges
• Primo cannot fetch contents by URLs and index it
− a plug-in based Exl approach should be developed to access eDOC
• Pushing data loaded only centrally into local Primo views
− we have 650.000 IVSCAN records loaded in ACC01
Solution
through local data processing - PPS
Local data processing – Primo data Preparation System (PPS)
11
What is PPS?
It is an application layer developed at OBVSG to process data produced by APM and eDOC enriching Central Primo data sources with additional elements.
APM – Aleph Publishing Mechanism, Exl
12
Online-Demo: Lokale Daten
Herausforderungen / Lösungen: Alle Daten für eigene Sicht
13
For a maximum of (parameterization) flexibility PPS produces a local copy of central MAB XML record for each Primo institution
• The institution applies its normalization rules and produces its own subset of _ data for Primo view
Central record ACC PPS
PRIMO Normalization
UBW PRIMO Normalization
ACC Copy
UBW HOL other
inst.
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW HOL UBW Bib. dataACC
HOL other inst.
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL other inst.
HOL UBW Bib. dataACC
HOL UBW HOL UBW Bib. dataACC
copy
Primo view ACC
Primo view UBW
14
Online-Demo: RTA
Herausforderungen / Lösungen: Real Time Availability - RTA
15
Primo uses system numbers of local bibl. records in Aleph Web X-service to check on-the-fly a title availability. The system number must exist in a local copy of MAB XML record produced by PPS
• PPS accesses Aleph local systems to obtain these data
• for a higher efficiency, the required data stored centrally and updated daily
System PPS
numbers
Bib.ID/SysNr
Aleph 500 UBW
Aleph 500 UBI
UBW MAB XML ACC MAB
after APMXML
Each ACC record is checked against local systems
16
Online-Demo: IVSCAN-Versorgung
Herausforderungen / Lösungen: IVSCAN Versorgung
17
IVSCAN articles to monograph works ~ 650.00 records linked to 38.000 head records but loaded only into the Central system. How to “push” these data to local systems in Primo?
• For each IVSCAN article PPS checks availability of its head record in a local system using the local file of system numbers (tag 599 is used)
• If the head record is found the article record is written into a local Primo data source
PPS
UBW MAB IVSCANXML ACC MAB
after APMXML
Check head records
UBI MAB IVSCANXML
IVSCAN record
System numbers
Bib.ID/SysNr
Write articles
18
Online-Demo: Abstract in Vollanzeige
Herausforderungen / Lösungen: Abstracts in Vollanzeige – „Plug In 1“
19
Using a customer developed plug-in program it is possible to manipulate parts of PNX record
• Currently OBVSG uses this approach to expand abstracts and integrate _“aggregate” objects from eDOC to the Primo full display
• The developed program accesses eDOC DB, fetches object and formats it as _necessary for the Primo display section of PNX record
Plug-in 1
PNX record ACC
fetches objects
eDOC Primo pipe
PNX record UBW
Herausforderungen / Lösungen: Volltext-Implementierung – „Plug In 2“
20
“Full text in Primo. It works now as we wanted but the way was long and winding”
• OBVSG decided for a solution when full texts are
transported into Primo not as a part of PNX record but via a separate import process
− the “import of pnx extensions” process runs outside of Primo pipes and requires a locally developed plug-in program
− each imported object is stored as a “pnx extension” record which is linked to a pnx record - these both parts are indexed
− the approach provides important decoupling of processing of bibliographic data and linked digital objects (both may have various update frequencies etc.)
Herausforderungen / Lösungen: Volltext-Implementierung – „Plug In 2“ (cont.)
21
• After development of plug-in program (10/2009), quite a lot of time and work was necessary to complete the planned implementation
− OBVSG was the first Exl customer implementing such approach
− a good co-operation with Exl developers brought functional improvements to the original approach
− many problems and bugs were solved and fixed
− a stable solution came from Exl in April 2010
Herausforderungen / Lösungen: Volltext-Implementierung – „Plug In 2“ (cont.)
22
Full text processing work flow as it is implemented on April 2010
Note. The process “Primo Pipe” runs for each data source (ACC,UBW etc.)
Produktionsablauf – Zusammenspiel der Komponenten
23
Überblick des täglichen Produktionsablaufs:
Produktion
22:00 23:30 23:50