VuFind

IAMSLIC Annual ConferenceBiennial EURASLIC Conference

Confluence of Ideas:Evolving to Meet the Challenges of Global Change

Brugge, Belgium1 October, 2009

Benoit PAUWELSUniversité Libre de Bruxelles (ULB)

Brussels

• Overview of functionality

• Technical overview

• VuFind @ ULB

Agenda

• VuFind is a library resource portal – designed and developed by Villanova University’s Falvey

Memorial Library– open source; web based– project leader: Andrew Nagy

• In the first place conceived as a new OPAC experience– Export/Import MARC records from ILS into VuFind– ILS specific drivers for holding information and availability

• VuFind can be used as a portal solution for all library resources– library catalog, article databases, digital library, institutional

repository & bibliography, and other library collections and resources

• http://www.vufind.org

VuFind

Customizable interface: localization and brandingUses Yahoo User Interface CSS, PHP Smarty templates, XML/XSL, AJAX, Javascript

ILS specific ‘drivers’ for holding information and availability status

Voyager, Aleph, III, SirsiDynix , Koha, Evergreen

Faceted search experience

“More like this” (algorithmically generated suggestions)

Save resource to organized lists

Browse

Internationalization

Social bookmarks

Zotero compatible (COINS)

Persistent URLs (bookmarkable searches)

Persistent URLs (bookmarkable searches)

Persistent URLs (bookmarkable searches)

RSS feed per search query

CiteThis, EmailThis, TextThis

Export record to Endnote, Refworks

Social features: tagging, comments

Enrichment: - book covers: Amazon, Syndetic, Google

Books - book item reviews: Amazon, Syndetic

• Linux, Solaris, Windows

• Uses Apache SOLR, an open source search engine based on Lucene

• Apache HTTP Server 2.2 or later

• PHP 5.1.0 or later, Smarty templates

• XSL, Javascript, CSS, YUI

• MySQL 4.1 or later

• Java J2SE JDK 1.5 or later

• GNU Aspell Library (optional spell checker)

Technical overview

• Export MARC records from ILS

• Use VuFind import facilty– Store each record in a separate file – Index in Lucene through SOLR web service

• Delivered SOLR schema.xml for MARC records

• Searching– VuFind communicates with Lucene through SOLR web

service– Returns unique identifier for records– Data of records fetched from filesystem (XSL)

Technical overview

• Keep VuFind functionality

• Don’t keep:– MARC based– Filesystem– Import utility; rather OAI harvest

VuFind @ ULB

Metadata

Harvester

Indexer

VuFind

OAI-PMH

MySQL

Metadata

JDBC

Lucene

Lucene doc

Samburu

SOLR

Metadata

Harvester

Indexer

VuFind

OAI-PMH

MySQL

Metadata

JDBC

Lucene

Lucene doc

Samburu

SOLR

Metadata exchange format

Compound Object DIDL / MODS (*)

DIDL/MODS XML record stored as-is in MySQL field

didl2solr.xsl transform DIDL/MODS XML record into Lucene document, according to specific schema.xml SOLR config file

<originalRecord>contains complete DIDL/MODS record

solr2brief.xslsolr2full.xsltransform <originalRecord> field into HTML

(*) http://homepages.ulb.ac.be/~bpauwels/NEEO/WP5/WP5 Technical guidelines.pdf/

• DI-fusion– Institutional library resources: IR,

etheses repo, digithèque, [ catalog, iconothèque ]

• BICTel– Regional e-theses repository

• Unicat– Belgian national virtual catalog

VuFind @ ULB