D4.2 (Modified)

Technical support documentation and help desk online
ECP-2007-DILI-517009
EuropeanaLocal
Technical support documentation and help

desk online
Deliverable number D4.2
Dissemination level PP
Delivery date 21 June 2009
Status Final
Stein Runar Bergheim, Olav Tuften, Idar

Thoresen Kvam, Rastislav Rehak with
Author(s)
inputs from EDL Foundation, and the
EuropeanaLocal project
eContentplus
This project is funded under the eContentplus programme1,
a multiannual Community programme to make digital content in Europe more accessible, usable and
exploitable.
1 OJ L 79, 24.3.2005, p. 1.
Technical support documentation and help desk EuropeanaLocal
online
Table of Contents
TABLE OF CONTENTS................................................................................................................................ 2
1 INTRODUCTION ................................................................................................................................. 4
2 DOCUMENT CONVENTIONS ............................................................................................................ 5
3 EUROPEANALOCAL GLOSSARY .................................................................................................... 6
4 OVERVIEW OF THE IMPLEMENTATION PROCESS ................................................................. 12
5 BEFORE YOU BEGIN ....................................................................................................................... 13

KNOWLEDGE REQUIREMENTS....................................................................................................................... 13
HARDWARE REQUIREMENTS ......................................................................................................................... 13
SOFTWARE/PLATFORM REQUIREMENTS ......................................................................................................... 13
6 SELECTING A REPOSITORY SOFTWARE PACKAGE ............................................................... 14
RECOMMENDATIONS ................................................................................................................................... 14
SOFTWARE COMPARISON MATRIX ................................................................................................................. 15
7 INSTALLING REPOSITORY SOFTWARE ON WINDOWS.......................................................... 18
WINDOWS REQUIREMENTS: .......................................................................................................................... 18
INSTALLING PKP HARVESTER ON WINDOWS ................................................................................................ 18
INSTALLING REPOX ON WINDOWS ................................................................................................................ 19
INSTALLING CELESTIAL ON WINDOWS.......................................................................................................... 20
8 INSTALLING REPOSITORY SOFTWARE ON LINUX ................................................................. 21
LINUX REQUIREMENTS ................................................................................................................................. 21
INSTALLING PKP HARVESTER ON LINUX ...................................................................................................... 22
INSTALLING REPOX ON LINUX...................................................................................................................... 23
INSTALLING CELESTIAL ON LINUX ............................................................................................................... 24
9 RELEVANT STANDARDS ................................................................................................................ 25
XML – EXTENSIBLE MARKUP LANGUAGE .................................................................................................... 25
DUBLIN CORE ............................................................................................................................................. 26
EUROPEANA SEMANTIC ELEMENTS .............................................................................................................. 26
THE OAI-PMH WEB SERVICE PROTOCOL ..................................................................................................... 28
10 METADATA EXTRACTION AND NORMALIZATION ................................................................. 31
METADATA EXTRACTION ............................................................................................................................. 31
METADATA FIELD MAPPING ........................................................................................................................ 31
METADATA NORMALIZATION ...................................................................................................................... 31
WRITING THE XML FORMAT ....................................................................................................................... 32
11 TESTING YOUR REPOSITORY ...................................................................................................... 34
12 TROUBLESHOOTING AND SUPPORT .......................................................................................... 35

EUROPEANALOCAL SUPPORT FORUM ........................................................................................................... 35
OTHER SUPPORT OPTIONS ............................................................................................................................ 35
BIBLIOGRAPHY ........................................................................................................................................ 37
TABLES AND DIAGRAMS ........................................................................................................................ 38
APPENDICES............................................................................................................................................... 39
2/39
online
APPENDIX 1: DUBLIN CORE METADATA ELEMENTS ...................................................................................... 39

APPENDIX 2: OAI PROTOCOL FOR METADATA HARVESTING .......................................................................... 39
APPENDIX 3: EUROPEANA SEMANTIC ELEMENTS V3.1 ................................................................................... 39
APPENDIX 4: ESE IMPLEMENTATION GUIDELINES ......................................................................................... 39
3/39
online
1 Introduction
This deliverable comprises the following technical support documents, together with an online
help desk which can be viewed at: http://www.avinet.no, all end-users will receive
username/password for this through the technical contact addresses provided in the technical
survey.
For test-access to the forum, please use the username “EuropeanaLocal” (without the quotes)
and password “europeana” (also without quotes and in lower-case letters).
This manual in combination with a series of technical workshops conducted throughout the
project are designed to support EuropeanaLocal content provider partners in establishing OAI-
PMH compliant repositories and populate them with metadata from their digital collections.
The manual can be used as a stand-alone reference for those who wish to take on the task of
implementing repositories without any aid, but will also be used as supporting material during
the regional Europeana Local technical workshops where content contributor partners will
receive hands-on training in installation, population and harvesting of OAI-PMH repositories.
4/39
online
2 Document Conventions
In order to ease the use of the manual, the following notations have been applied consistently
throughout the document:
Examples including script code and data sections are shown in Courier
typeface on grey background
Italic text should be read and followed carefully as such tasks require careful attention.
Tips and tricks which may be useful for content contributors will be highlighted in bold.
5/39
online
3 EuropeanaLocal Glossary
The following list defines useful terms and abbreviations for content contributors to
Europeana. The definitions below are valid when used within the context of Europeana Local.
Several terms have multiple definitions, the definition given below is not necessarily the official
one.
Term Definition
Application profile Adaption of a metadata standard for a specific purpose or

use.
Athena A sister-project to EuropeanaLocal seeking to develop

tools and methods for improved metadata extraction and
enrichment from MLA content contributors across Europe.
Authentication In the context of Europeana Local, the process of verifying

the identity of an external party who wants to harvest a
source repository.
Authority file An authority file defines a list of known unique entities.

Example:
 A list of person names
 A list of geographical names
 A list of time periods
Creative Commons A set of predefined rights expressions which may be used to

state the IPR associated with content from a local content
contributor.
Dictionary A list with definitions of words and/or translations of words

between different languages
Digital collection In the context of Europeana Local, a digital collection is a

uniform database held by a content provider.
Dublin Core Metadata standard, defines a wide range of descriptive

metadata for any kind of item
Dublin Core (qualified/simple) Dublin Core with the minimal set of mandatory fields
present and populated.
EDLF Abbreviation for the European Digital Library Foundation.

Foundation which among other things is responsible for
running and maintaining the Europeana service.
ESE Abbreviation: Europeana Semantic Elements, see

Europeana Semantic Elements
6/39
online
Europeana A web application run and developed by the EDLF. The

equivalent of Google for searching within the domain of
quality controlled cultural heritage information sourced
from MLA organizations across Europe.
EuropeanaLocal This project, aiming to identify the shortest path for local or
regional content providers across Europe to connect their
content to Europeana.
Europeana Semantic Elements An application profile of Dublin Core adapted by the

European Digital Library Foundation for the purpose of
enabling desired functionality in Europeana.
Europeana themes A set of themes proposed by Europeana which should be

guiding for which collections should be prioritized.
Available off the Europeana website but is subject to
change.
GIS Abbreviation for Geographical Information System
Harvesting See metadata harvesting
Harvesting authority A registry contained in a harvesting repository (e.g.

Europeana itself, national aggregators, etc) where any
number of source repositories may be entered and
subsequently harvested.
ID Abbreviation for identifier. See identifier, persistent

identifier
Identifier A number or string which is used to uniquely identify a

content item within a global or local scope.
Example: an item might have the ID 23568 in a local

collection but once this item is harvested by Europeana
other items with the same ID may exist because the ID is
unique only unique within the scope of the local collection.
By prefixing a local ID with a unique string it will be unique

in the global scope. A common way of doing this is
through the use of URIs. The URI may be the web address
(URL) which if typed into a web browser will show the
presentation page of the item from the web site of the
content provider.
IPR Abbreviation for Intellectual Property Rights
Item An Item is one record of metadata describing an abstract or

physical entity in terms of the elements described in ESE.
7/39
online
See: ESE, application profile
Metadata Data describing a resource in terms of intellectual property

rights, what, where, when and who. In the context of
Europeana Local, the resources described have their own
unique web page.
Metadata enrichment The process of improving the quality of metadata through

manual or automated methods.
Metadata extraction The process of reading metadata from custom digital

collections held in e.g. SQL-databases and writing them in a
format suitable for import to a repository.
Metadata harvesting The process of reading metadata from a source repository

over the web, typically using the OAI-PMH protocol which
defines available web service methods (verbs) and their
response format.
Metadata mapping The process of mapping the fields of a source metadata

format to the fields of a target metadata format. This
process may include splitting of fields which may require
extensive manual quality control. It may also include
merging of fields and the automatic addition of static
metadata which are missing in the source metadata format.
Miniature In the context of Europeana, a miniature is a simplified

representation of a content item, e.g. a thumbnail edition of
a photo, a short audio-sample of a music track or a preview
clip of a video file.
MLA Abbreviation for Museums, Libraries and Archives.
Multi-lingual metadata Metadata in more than one language. While this term is
quite simple it points to a rather important problem. What
can we use the metadata for if they are not in a common
language?
If the source vocabularies are not mapped to a common,

multi-lingual vocabulary, searches for language specific
words like “chair” will render hits only for English language
resources while the equally important French “chaise” will
be omitted.
OAI Abbreviation for Open Archive Initiative. A body

developing standards for storage, exchange and access to
MLA data.
OAI-ORE Abbreviation for Open Archive Initiative – Object Reuse
8/39
online
and Embedding. A standard for harvesting of rich content

items (images, audio-files, texts) as well as their metadata.
OAI-ORE is envisaged to enable sophisticated services at
the harvesting end but requires more efforts from content
contributors to populate source repositories and requires
larger bandwidth as the volume of data transferred will be
large.
OAI-PMH Abbreviation for Open Archive Initiative – Protocol for

Metadata Harvesting. A standard which defines the allowed
calls to and response from a web service offered by a
repository. Is used to harvest the content of local
repositories into aggregator repositories such as regional,
national or trans-national portals such as Europeana itself.
Object An Object is a composite representation of digital content

consisting not only of metadata but also the actual content.
An Object would typically be represented in a format such
as RDF and contain embedded images, text, audio, video
corresponding to the actual content resource.
Ontology A comprehensive set of concepts with relationships

between them. Ontology is defined in a language such as
OWL or RDF.
Example:
A definition of a concept in an ontology typically consists
of a source-relationship-target triplets as shown below:
1. Source: A chair
2. Relationship: Has parts
3. Target: A seat
An ontology is instrumental in creating intelligent search

and retrieval services as it may be used to aid users in
searching across collections which are using differently
named concepts to describe their containing items.
OWL Abbreviation for Ontology Web Language. An XML-based

format for defining ontology.
Persistent identifier An identifier which remains unchanged over time.
RDF Abbreviation for Resource Description Framework. An

XML based format used to described metadata for content
items or embed object data. RDF has a rich model for
defining relationships between items. RDF may also be used
to express ontology.
Repository Metadata storage software which enables harvesting

authorities to connect to it remotely over the Internet and
9/39
online
download it‟s content in one or more formats. Repository

software typically consists of a database with a web
application which allows import of data, searching and
browsing of the metadata and an OAI-PMH compliant web
service interface. Examples of repository software include
Fedora, Greenstone and DSpace.
Schema (database) A database schema is a space in a database where

information is kept separate from other information in the
same database. Tables in a schema share some properties.
Schema (metadata) A metadata schema is the same as a metadata format. ESE

is a metadata schema based on Dublin Core. See:
application profile.
Semantic Web A type of web service which allows “intelligent” searches

for information through definition of relationships between
rich content items or linkage to a common ontology.
SKOS XML format for describing a vocabulary, used to facilitate

mapping between different vocabularies.
SKOSification/SKOSifying The process of implementing a local vocabulary in the

SKOS XML-format
Surrogate The representation you can make of a metadata item based

on the data contained in the metadata itself. A surrogate
would typically be the page showing all the details of an
item returned by a search in Europeana.
Thesaurus Plural: thesauri. A dictionary defining synonyms for words.

May be used to automate the mapping from one vocabulary
to another.
URI Abbreviation for Uniform Resource Identifier. World Wide

Web Consortium defined standard for uniquely identifying a
resource on the Internet.
Example:
http://www.europeanalocal.eu/id=12
Vocabulary A vocabulary is a list of words systematically used to

describe content in a digital collection.
Vocabulary mapping The process of mapping the “words” of one vocabulary to

the “words” of another. May be automated through the use
of thesauri and/or dictionaries which automatically define
relationships between items based on synonyms.
10/39
online
WHAT metadata Descriptive metadata indicating the thematic nature of the

content through a classification system, vocabulary,
dictionary or free text keywords. Example: image, people,
black and white, hunt, fox, horses
WHEN metadata Metadata indicating the temporal validity of an item.

Example:
 14.08.1942
 1942
 The 1940s
 Second world-war
WHERE metadata Metadata indicating the geographical location of an item.

Example:
 coordinates
 geographical names
WHO metadata Metadata indicating known named entities which are

relevant to an item. If the item is a photo, this may be the
name of people in the photo. If the item is an article it may
be the name of an organization mentioned in the article.
11/39
online
4 Overview of the Implementation Process
The detailed procedures described in this manual will take you through the seven steps
required to enable your locally held content to be harvested by and shown in the Europeana
service.
Once you have established a repository, it may be harvested by any number of harvesting
authorities dependant on who you give access to your repository. Intelligent semantic search
services may be built based on these aggregated services.
1. Establish compliant hardware/software platform
2. Install repository software
3. Configure repository software
4. Extract and normalize metadata
5. Populate repository
6. Test that harvesting is working
12/39
online
5 Before You Begin
This section describes the prerequisites for carrying out the tasks in this manual. There are
three sets of requirements which need to be met:
Knowledge requirements
It is assumed that users of this manual will be technically skilled and will have advanced
understanding of PCs. Users should be experienced in regular software installation. Users do
not need to have programming expertise but having played around with some simple scripts
will be an advantage.
Hardware requirements
A server which should be usable as a repository needs to meet the following minimum
requirements to deliver satisfactory performance:
 Processor speed: >= 1 x 2,0 GHz
 Storage space: dimensioned to data volume >= 200 GB
 Physical memory: >= 2GB
Software/platform requirements
Servers need to be installed with basic operating system software, either Linux/Unix or
Windows based. The following software requirements needs to be met:
 Ubuntu Linux
 Apache Web Server
 Tomcat Application Server
Or
 Windows 2000/2003 Server
 Apache Web Server / Internet Information Server
 Tomcat Application Server
13/39
online
6 Selecting a Repository Software Package
Repository software serves several purposes. It stores metadata and can be used as a basis for
developing custom local services. It typically allows browsing of metadata through a web
interface and it offers the data to harvesting repositories over the Internet through TCP/IP
communication.
The main requirement for content to be harvestable by Europeana (and regional/national
aggregators) is that the XML returned by issuing a call to the repository is compliant with the
OAI-PMH standard.
Repositories may be classified into static repositories and dynamic repositories.
 A static repository is merely dump of metadata from a digital collection into a static XML-file.
This file can then be harvested by aggregating repositories over the Internet
 A dynamic repository returns OAI-PMH compliant XML as well as support a set of OAI-PMH
requests and responses.
A number of different software packages offer dynamic repository functionality and if

technically sophisticated partners wish to contribute content to Europeana she or he is free to
do so by installing any compliant product. Partners who are in need of technical aid in order to
set up their repositories are however recommended to choose one of the software packages
which are described in this manual.
Recommendations
There is a large number of available repository software packages, most of which are open
source and claim to be superior to each other left, right and center. EuropeanaLocal technical
partners have therefore performed an assessment of different repository software packages
based on a set of simple criteria reflecting the needs of local and regional content providers.
6.1.1 Recommended content provider repository software

For content providers, EuropeanaLocal technical partners have seen two criteria as being key:
 Ease of installation
 Ease of running and maintenance
Based on these criteria EuropeanaLocal technical partners recommend Repox (TEL+ Project,
2009) as content provider repository. The software package is easy to install, comes with its
own web server and database and supports a multitude of ways of ingesting content from local
files. The software also has an attractive interface for defining crosswalks between different
metadata formats and is being further developed within the Europeana family of projects
through Europeana v1. Repox is scheduled to support ESE starting in October 2009 until
which time such support will have to be customized.
Other good choices include Celestial (simple) and Fedora (comprehensive).
6.1.2 Recommended aggregator repository software

For aggregators, EuropeanaLocal technical partners have seen three criteria as key:
 Ease of installation
 Ease of running and maintenance
 Customizability
14/39
online
The final criterion is defined because aggregators are not merely there to aggregate content to
be passed on to Europeana but mainly to build value-added search and discovery end-user-
services at regional and national level. The ease of developing such services based on the
aggregated content is therefore an important factor in selecting a repository software package
for an aggregator.
Based on the above criteria, EuropeanaLocal technical partners recommend PKP Harvester
(Public Knowledge Project, 2009) as an aggregator repository. PKP Harvester is a very mature
harvesting software package with a comprehensive APi for programming your own
applications based on its data store. Recently a promising OAI-PMH provider interface has
been added which makes this a good choice for aggregators wishing to develop their own
applications as well as providing content for Europeana.
Other good choices include: Repox (provided all features promised will be present in the next
version – currently lacks harvester), Fedora (provided additional harvester software is
integrated or ingestion of content from underlying entities does not happen according over the
OAI-PMH protocol. Austria has also shown good results using Celestial though this software
package does not offer any API for customization or development of specialized search
applications.
Software comparison matrix
Table 1: Feature comparison matrix for assessed repository software packages
Issue Description Celestial Repox PKP Fedora1 Static

Import n y n y n
Harvester Has built-in y y y n n
Sets y y y y n
- harvest y y y y
- provide y y y y
- preserve y n n y
Flow control Resumption y y y y n
Provenance Identifier and y n n y n
rights
Stores format the actual tel the actual fobject the
act.
Crosswalks n y y y n
- existing Repox will get na dc, marc dc, marc, dc, marc,
ESE mods mods,
transformation mets
Search Fedora has RDF- n n y y n
based
ResourceIndex
Database Mysql Derby Mysql Mysql xml
Postgres Mysql Postgres Postgres file
Oracle
- how stored As xml, fields, xml xml partly both xml
both
Ease of Demanding e e e n e
installation Normal Windows
Easy n/d Linux
1 Note. Fedora is also representing the tree other big content systems, DSpace, Eprints and Greenstone. They
would give much the same results in the table.
15/39
online
OS Platform Linux/Windows both both both both any

Web Server Apache Jetty Apache Tomcat Perl
IIS IIS
Language Programming Perl Java PHP Java Perl
language
ESE support Must make n/a Cross- Plug-in Cross- n/a
crosswalks. walk walk
Repox will
support ESE.
Capacity Recognised - - - 30 mill 5.000
Provider databases
Capacity 10.000 records at 1.5 min 1.5 min 3/18 min 14 min n/a
Harvester our test site, PKP
search off/on
Scalability Recognised y y y y n
databases.
Support/ Repox is n y y y n
Maintenance maintained in
eContent
projects.
Deleted records Repox will y n n y n/a
support.
Is reported to
PKP.
Suitable for Provider A,P P A,P A,P P
whom Aggregator
Installations Registries like Some TELPlus Many as Many
Roar, Oaister, 1180 Other? harvester
Celestial shows
Native language English is n n y y n
standard
Supporting tools
and utilities
Customizability Demanding D D N,E N,E
Normal
Easy
API’s n N y y n
Security Simple, more s S a a n
Advanced
Admin GUI very Simple simple simple none
simple
Command line Harvest/Import y N y y y
Content n N n y n
management
While the above matrix shows the comparison of features between the different software
versions, the below table shows what are the major strong and weak points of the assessed
technologies in the context of EuropeanaLocal.
Table 2: Advantages and drawbacks of repository software packages
Tools Pros Cons

Static - Simple, just a XML file - No sets.
repository - General, replaces import - No resumption token.
- Only few records.
Celestial - Perl, fast - No Import
- Simple, but good datamodel - No Search
- Mirror, preserve everything - Too simple admin
- Takes any metadata format - Lacks crosswalk to dc, must store it
16/39
online
- Has provenance - No maintenance

Repox - Java, pretty fast - No search
- Will get ESE support and Import - No command line
- Maintained - Memory consuming, Static a 100 recs
- By eContent project Is reported:
- Don’t handle deleted record
- Overwrite sets
- Must be harvested by set
PKP - PHP, fast when Search disabled. - No Import
Harvester - Flexible, several API Is reported:
- Supported - Deletes ‘deleted record’
- Maintained - Overwrites sets
- Search engine - Cannot be harvested by set
- New identifier, preserve old for search
Fedora - Java - No harvester
- Scalable - Content system, extended
- Several APIs, e.g. Search
- Maintained
17/39
online
7 Installing Repository Software on Windows
We have tested our repositories on both work station and server computers. They all work fine
on Windows 2000/XP/2003 servers.
In this setup we have been using Apache as web-server on the windows platform also, this is
mainly because we prefer to keep the two environments as identical as possible. However
things will also work with IIS on Windows.
Windows requirements:
 Windows 2000/2003 Server or newer

 PHP and Perl needs to be installed, (works equally well on IIS as on Apache)
 If Repox is chosen, it will install it’s own Java Servlet engine, Jetty.
 If Celestial is chosen, Avinet will supply a modified package of this software
 PKP Harvester is a PHP product and works as long as PHP is installed
 A database.
o Repox comes default with Derby, a java based RDB.
o Celestial should work on most databases which support the use of blob or blob-like
data types. We have been using mySql.
o PKP Harvester supports numerous databases as long as the php-driver files are
installed.
o Regarding PKP and PHP. Make sure to enable the following extension in the PHP.ini
file: extension=php_mbstring.dll before you import any data.
 Make sure the computer can access the internet. If this is sensitive data, one might want to
setup some type of authentication.
7.1.1 Installation details

Download and install the latest version of Perl. Perl is needed regardless of what repository
you choose.
Perl can be found at www.perl.net. Installation should be pretty straight forward. When perl is
installed you should have access to cpan. Cpan is a tool for installing additional perl libraries.
You need to atleast install the following ones:
Download and install a database. Repox have its own, Derby. Celestial have been tried and
tested on mySql and PKP Harvester supports 8 or 9 different ones. We have used mySql and
PostgreSQL for PKP Harvester installations.
You do not need to download Apache webserver for Windows. PKP Harvester and Celestial
should run on IIS. Repox has its own java server and can run on that.
If you are going to use Repox, download and install the latest java version. From
http://java.sun.com/javase/downloads/index.jsp
Installing PKP Harvester on Windows
Download and extract the files into the c:\inetpub\wwwroot directory. Make sure that the
cache- and public directory is writeable. Also make sure the config.inc.php file is writeable.
Make sure the php-files for your chosen database are installed.
7.1.2 Configuring MySQL

Log into the database, create a user and set a password for it.
18/39
online
Mysql –u root.
After logging in it would be smart to give the rootuser a password.

Set password = password(’something’);
Create a database user for PKP Harvester named harvester2 :

create user harvester2@localhost;
set password for harvester2@localhost = password(’harvester2’)..
7.1.3 Configuring PKP Harvester

After completing the above steps, open your web browser, go to http://localhost/harvester2
and configure the following options:
 Locale settings:
o Primary Locale English
o Client Character set Unicode (UTF-8)
o Connection character set Unicode(UTF-8)
o Database character set Unicode(UTF-8)
 Security settings
o SHA1, since it is more secure then MD5
 Administrator Account
o Type in the username/password you created when you set up the database’
o Insert an administrator e-mail address
 Database settings.
o If there are brackets around the database you wanna use, the php drivers for that
database has not been installed.
o Host is usually localhost
o Insert the username/password you created in the database setup
Push the Install Harvester2 button. The installation process usually takes no longer than 3-4
seconds.
If there are any errors in the data you have inserted you will get the same installation page
again with an error message.
If everything looks ok you will get a confirmation page telling you that the installation was
successful and a link to the login page.
Login and enjoy
Installing Repox on Windows
Download the installer from http://repox.ist.utl.pt/REPOX_1.2-installer.jar and install it by

double clicking the downloaded file.
Follow the on-screen instructions. Make sure the user making the installation has got
administrator privileges on the computer where the installation takes place.
On Windows Vista you have to do the following:
Go to Start > All Programs > Accessories > then right-click on "Command Prompt" and select "Run as
administrator".
Then go to the directory where you put the file and run
19/39
online
java -jar install.jar
When installation is done, login – and enjoy.

Installing Celestial on Windows
Download the package containing the Celestial files. A modified version is made available by
EuropeanaLocal.
With the files we have supplied some libraries that need to be built and installed. Make sure
your system has a make-builder installed. One version can be found here
http://unxutils.sourceforge.net/
Unpack the library file and do the following
perl Makefile.pl
Make
make install
When that is done, there are some other PERL libraries that needs to be installed, if they are
not already present:
XML::Parser::Perlsax, XML::Parser, XML::DOM, XML::Writer
Make sure the user and the table name have been created already in your database of choice.
7.1.4 Configuring MySQL

We assume mySql is being used:
mySql –u root
When prompted, type in the root password for MySQL determined at time of installation.
Create user celestial@localhost
Set password for celestial@localhost = password(’password’);
Create table celestial
Grant all privilieges on celestial.* to celestial@localhost
When this is done, run the installer script

perl –w install.pl, and follow the onscreen instructions.
After installation. Run http://localhost/celestial/config and set it up to harvest from files or

from other repositories.
20/39
online
8 Installing Repository Software on Linux
We have tested all our solutions on both desktop and server versions of Ubuntu linux. And the
solutions works fine on either platform.
On both versions, several software packages needs to be added after a clean installation. If you
use an existing server, chances are that many of those packages are already installed.
We have used Apache as web-server for PKP Harvester and Celestial, alongside the built-in
Jetty web server that comes with Repox.
As database we have used MySQL for PKP Harvester and Derby which comes packaged with
Repox.
Linux requirements
Linux distribution of choice

PHP and Perl needs to be installed, along with a web server
If Repox is chosen, it will install its own Java Servlet engine, Jetty
If Celestial is chosen, EuropeanaLocal will supply a modified package of this software
PKP Harvester is a PHP product and works as long as PHP is installed
You need a database. Repox comes default with Derby, but should work with other RDB‟s,
like PostgreSQL, MySql, MS SQL Server
Celestial should work on most databases that support the use of blob or blob-like data types.
PKP Harvester supports numerous databases as long as the php-driver files are installed.
One little thing regarding PKP and PHP. Make sure to install the php extension for mbstring before
you import any data. E.g:
apt-get install php5_mbstring
Make sure the computer has accessible to the internet. If this is sensitive data, one might want
to setup some type of authentication.
8.1.1 Installation instructions on Ubuntu platform

Start with a clean ubuntu installation
Type in the nessecary information along the process, language, keyboard, server name, user
name and password.
When you get to choose what type of server you would like to install, choose only the ssh-
server option.
When all is done it‟s time to log on with the user you created.
First of all, start with an update.
Sudo apt-get update
Sudo apt-get upgrade.
Then we start to install necessary software:

Sudo apt-get install apache2. This is the webeserver.
Sudo apt-get install php5-cgi Cgi binary for php
Sudo apt-get install perl (Is part of the ubuntu installation)
Sudo apt-get install sun-java6-sdk. Java is only nessecary if you want to
run Repox as repository.
21/39
online
Install your favourite database for the repository. The solutions we have tried out uses MySql
and PostgreSQL. Repox comes with its own RDBMS, Derby. PKP Harvester supports a
multitude of databases.
Sudo apt-get install mysql
Install the php databasedriver for your selected database, like php5-mysql.
Sudo apt-get install php5-mysql
Log on to the database and create a user, and set a password.

Mysql –u root
After logging in it would be smart to give the rootuser a password.

Set password = password(’something’);
Create a user for PKP Harvester:

create user harvester2@localhost
set password for harvester2@localhost = password(’harvester2’)
Create a user for Celestial:

create user celestial@localhost
set password for celestial@localhost = password(’celestial’).
Now we are ready to install the repositories.

Installing PKP Harvester on Linux
Download the source from http://pkp.sfu.ca/harvester2/download/harvester-2.3.0.tar.gz

Unzip/untar the file and put the contents in your www-directory
C:\inetpub\wwwroot or /var/www or similar
Make the cache and public area writeable to anyone. To make installation easier also set all
rights for everyone on the config.inc-php file. And change it back to read only afterwards.
If you choose to use something else than ubuntu, make sure that the php5_mbstring file is
installed.
Start your favorite web browser and go to http://localhost/harvester2
On the Pre-Installation Steps screen - make sure all the lines say “yes”.
Configure locale settings

 Primary Locale English
 Client Character set Unicode (UTF-8)
 Connection character set Unicode(UTF-8)
 Database character set Unicode(UTF-8)
Configure Security settings

 SHA1, since it is more secure then MD5
Configure Administrator Account

 Type in the username/password you created when you set up the database
 Insert an administrator e-mail address
Configure Database settings.

 If there are brackets around the database you wanna use, the php drivers for that database
has not been installed.
22/39
online
 Host is usually localhost

 Insert the username/password you created in the database setup
Push the Install Harvester2 button. The installation process usually takes no longer than 3-4
seconds.
If there is any errors in the data you have inserted you will get the same installation page again
with an error message.
If everything is ok you will get a confirmation page that the installation was successful, and a
link to the login page.
Login and enjoy
Installing Repox on Linux
The previous versions of Repox needed access to an x-server in order to be installed. But in the
newest relase, an easier way of installing it on Linux has been made.
We will support you with an altered version of Repox which works. This is due to some small
alterations that most likely has to be done in the installation script.
Unzip the file REPOX_1.2_unix-installer.zip. You‟ll end up with a directory named
repoxlinux. Move this to the /opt catalogue.
Run the install.sh script and follow the onscreen instructions.
Urn:name.of.your.choice
Jetty server port – keep the 8080 port
Data path /opt/repoxlinux
Admin email
User login – keep the admin
User password – whatever you like. Just remember it
When the installation is done, Repox will be available on http://localhost:8080 unless you
specified something different during installation.
If Repox is the only web application that is going to run on your server, you don‟t need apache
webserver. The Jetty server installed with Repox should be sufficient.
If you are planning to run other web-related material on this server, apache is a must. Then we can
”hide” Repox behind the apache-server, and we can access Repox on the standard html port, 80.
To do so we need to make some changes in the httpd.conf
LoadModule proxy_module /path/to/modules/mod_proxy.so

ProxyRequests Off
<proxy *>
Order deny,allow
Allow from all
</proxy>
ProxyPass /repox http://localhost:8080/repox
ProxyPassReverse /repox http://localhost:8080/repox
ProxyPreserveHost On
Proxystatus On
23/39
online
Installing Celestial on Linux
The original Celestial distribution has been slightly modified by Technical Partners. The modified
package is available to partners.
Download the file celestial.tar.gz

Untar the file:
tar xzf celestial.tar.gz
You will end up with X directories. Move the contents of the bin directory to your cgi-bin
catalogue and the data directory to /opt/celestial.
The config file is pre-setup for use with mysql
Included in this package is another tarfile, called oai-perl-2.15.tar.gz. It contains perl libraries
that provides an API to OAI-PMH.
Tar xzf oai-perl-2.15.tar.gz
Cd into the directory that was created, build a makefile with the command:
perl Makefile.pl
Then run make

Then run make install
With the libraries installed, we can turn to the database.
perl –w install.pl
And follow the onscreen instructions.

After installation. Run http://localhost/celestial/config and set it up to harvest from file or from
another repository.
24/39
online
9 Relevant Standards
The following chapter describes relevant standards for the implementation of content provider
repositories among EuropeanaLocal content providers across Europe.
XML – eXtensible Markup Language
XML is at the heart of all data exchange in EuropeanaLocal, it is therefore imperative that
technical personnel are acquainted with this way of embedding data before attempting to create
any extraction scripts and/or ingest data into a repository.
XML is a very simple and flexible markup language and it suffices to be aware of the following
conventions and concepts in order to create standards compliant XML. Consider the below
example.
<?xml version=”1.0” encoding=”UTF-8”?>
<root>
<namespace:element attribute=”attribute value”>element value</element>
</root>

Let us take a look at the first line:

This line looks like any HTML tag, except inside the tag follows a question mark <?xml … ?>.
This signifies the start of an XML formatted text document.
The important thing to be aware of here is the encoding attribute which in the above example is
using UTF-8 – a character set capable of storing multi-byte strings, thereby enabling the
representation of any special characters found outside the Latin alphabet without breaking. If
you input a character into the XML which is not supported by the character encoding given in
this line, an error will result.
XML is not tolerant to characters like >, <, & or similar. For this reason, these should be
encoded into their corresponding XML-entities during export in order to avoid broken
characters in exported files. The entity “&” can for an example be represented legally like this
“&”.
Let‟s move on to the next line:
<root>
An element is a keyword enclosed by angular brackets “<>”. A start element may look like
this: <element> and an end-element may look like this: “</element>” with the containing
value between them. Empty elements may look like this: “<element/>”.
The first element following the initial xml-tag, is called the “root element”. Any subsequent
element is known merely as elements.
Continuing to the third line we find the following:
<namespace:element attribute=”attribute value”>element value</element>
In this line we see that the element keyword contains a colon. The keyword before the colon
indicates the “namespace” to which the keyword after the colon belongs. A namespace is a
controlled set of element names associated with a “schema” which defines permissible values
and attributes for the each element.
In addition to the namespace:keyword inside the element tag, we find another keyword
followed by an equal sign and a set of double quotes encapsulating a value. This second
25/39
online
keyword is called an “attribute”. Attributes may look like this: „attribute=”attribute value”‟
and will always exist within elements.
Finally, let us have a look at the notation for adding inline comments to XML markup.

Any text between a “<!—“ and “-->” will be considered as comments and ignored by XML
interpreters. This is handy if we wish to include additional attributes which are not recognized
by standards such as ESE 3.1 and Simple Dublin Core.
There is a wide range of good XML-references on the Internet, among which EuropeanaLocal
technical partners would recommend w3schools XML-tutorial, easily found using Google. The
above will suffice as a basic primer to prepare a basic XML-file. (w3schools.com, 2009)
Dublin Core
Dublin Core describes standardized ways of structuring and naming metadata elements so that
they can be read and interpreted properly by Dublin Core aware computer systems.
The following example shows a Simple Dublin Core metadata description of the resource
found at http://www.europeanalocal.eu, the project web site of EuropeanaLocal.
<?xml version="1.0"?>
<metadata
xmlns:xsi=”http://www.w3.org/2001/XMLSchema-instance”
xmlns:dc=”http://purl.org/dc/elements/1.1/”>
<dc:title>
EuropeanaLocal Project Web Site
</dc:title>
<dc:description>
EuropeanaLocal is a project funded under eContentplus, identifying
best practices for inclusion of locally held content into Europeana.
</dc:description>
<dc:publisher>
EuropeanaLocal
</dc:publisher>
<dc:identifier>
http://www.europeanalocal.eu
</dc:identifier>
</metadata>
This is relevant to EuropeanaLocal because the core metadata set determined by Europeana is
an extension of Dublin Core by the name of Europeana Semantic Elements, described in
greater detail below.
The Dublin Core metadata set is defined and evolved by the Dublin Core Metadata Initiative
and its working groups. A lot of useful information as available at the Dublin Core Metadata
Initiative‟s web site including their Guidelines for implementing Dublin Core in XML
(Powel & Johnston, 2003)
Europeana Semantic Elements
Europeana Semantic Elements is an application profile based on Dublin Core. It defines a

number of properties which can be used to describe digital objects. These properties are listed
26/39
online
in the table below. Please notice the XML concepts namespace and element. The column
occurrence states whether the element is optional (0..1/0..*) mandatory (1..1/1..*) and whether
it can occur only once or multiple times in the XML (1/*)
Table 3: ESE elements
Namespace Element Element Occurrence

(Abstract) (Refinement)
Dc Title 1..1
dc:terms alternative 0..*
Dc Creator 0..*
Dc Subject 0..*
Dc Description 0..*
dc:terms tableOfContents 0..*
Dc Publisher 0..*
Dc Contributor 0..*
Dc date 0..*
dc:terms created; 0..*
dc:terms issued 0..*
Dc type 0..*
Dc Format 0..*
dc:terms extent; 0..*
dc:terms medium 0..*
Dc Identifier 1..1
Dc source 0..*
Dc Language 0..*
Dc relation n/a
dc:terms isVersionOf; 0..*
dc:terms hasVersion; 0..*
dc:terms isReplacedBy; 0..*
dc:terms replaces; 0..*
dc:terms isRequiredBy; 0..*
dc:terms requires; 0..*
dc:terms isPartOf; 0..*
dc:terms hasPart; 0..*
dc:terms isReferencedBy; 0..*
dc:terms references; 0..*
dc:terms isFormatOf; 0..*
dc:terms hasFormat; 0..*
dc:terms conformsTo 0..*
Europeana relation n/a
europeana isShownBy1; 0..1
Europeana isShownAt 0..1
Dc Coverage 0..*
dc:terms spatial; 0..*
dc:terms temporal 0..*
Dc rights 0..*
1 Either isShownAt or isShownBy MUST be present in the XML

27/39
online
Dc terms n/a
dc:terms Provenance 0..*
Europeana userTag1 0..0
Europeana unstored2 0..1
Europeana object3 0..1
Europeana language 1..1
Europeana provider4 1..1
Europeana type5 1..1
Europeana uri6 1..1
Europeana year7 0..1
Europeana hasObject 1..1
Europeana country 1..1
In addition to the elements, Europeana Semantic Elements also define an attribute which can
be used to specify the language of the element values, xml:lang. This attribute can be used for
all text-based elements with multiple occurrences to state whether the language of the element
value is e.g. English (en), French (fr) or German (de).
<dc:subject xml:lang=”en”>weather</dc:subject>
<dc:subject xml:lang=”de”>wetter</dc:subject>
The usage and content of Europeana Semantic Elements is described in greater detail in the
two documents Specification for the Europeana Semantic Elements (Clayphan, et al.,
02/2009) and Metadata Mapping & Normalisation Guidelines for the Europeana
Prototype (Clayphan, Lupovici, Sugimoto, & Verleyen, 05/2009). Both of these documents
are attached as appendices to the implementation manual, see Appendix 3: Europeana
Semantic Elements v3.1 and Appendix 4: ESE Implementation Guidelines .
The OAI-PMH web service protocol
While the standards we have been looking at above are related to the metadata content and
how it should be formatted, OAI-PMH provides us with a means of transporting metadata
from an OAI-PMH compliant provider repository to an OAI-PMH aggregator repository by
means of harvesting.
OAI-PMH is a simple web service protocol which can be invoked calling the URL of an
installed repository with one out of six arguments – called verbs. Each of these will return a
specifically formatted XML string to the requestor web service, making it possible to interpret
the nature and quantity of the content held in the provider repository as well as downloading
its metadata content. The six verbs are as follows:
1 This element is only meant to store user tags contribute through the Europeana application and does not need
to be included in the content providers‟ XML
2 Include any metadata properties which doesn‟t fit into the other elements but which may be interesting to
include for indexing purposes in the “unstored” element.
3 The object element will be populated by Europeana if left out from the XML
4 The provider element will be populated by Europeana if left out from the XML
5 One of TEXT, IMAGE, SOUND or VIDEO, Europeana controlled vocabulary
6 Will be added automatically by Europeana
7 Four digit year. Currently no support for dates before year 0. Planned to be implemented shortly.
28/39
online
1. Identify
2. ListMetadataFormats
3. ListSets
4. ListIdentifier
5. ListRecords
6. GetRecord
The most important of these verbs is “ListRecords” as this is the one returning actual metadata
records from the source repository. This is the verb requested by harvesters who connect to
source repositories, EuropeanaLocal content providers must therefore make sure that the
response is properly formatted in accordance with the XML format rules as shown below.
<OAI-PMH>
<ListRecords>
<record>
<header>
<identifier>
<dateStamp>
<setSpec>
</header>
<metadata>
<!—ESE metadata elements here -->
</metadata>
<about>
<rights>
<provenance>
</about>
</record>
</ListRecords>
</OAI-PMH>
Below you will find a description of the most important elements of the OAI-PMH
ListRecords XML format.
9.1.1 Identifier
To be able to separate between updated and added records during incremental harvesting, it is
necessary that a unique identifier for each item exists in the provider repository.
OAI-PMH allows for identifiers which are not directly related to physical locations, but instead
uses a system of locally unique identifiers within a repository. The format of the unique
identifier must correspond to that of the URI (Uniform Resource Identifier) syntax. The syntax
of an identifier is:
oai-identifier = scheme : namespace-identifier : local-identifier
The scheme is always the literal string “oai”. The namespace-identifier is usually related to the
domain-name of the organization hosting the repository.
The local-identifier is some identifier which is unique in the context of this particular
repository. An example of an identifier from ArXiv.org might look like this: oai:arXiv.org:hep-
th/9901001.
29/39
online
Globally unique identifiers are useful in that they make it possible to trace a metadata record
back to its source repository, and because they can be used to build rich and interlinked
services based on metadata harvested from different repositories.
9.1.2 Timestamps
The from/until parameters of the harvesting verb ListRecords is related to the dateStamp
element with granularity of day or second.
9.1.3 Sets
Sets are an optional feature of OAI-PMH offering setSpec in the header section to divide the
records reflecting e.g. the organizational or thematic structure of the records. As these are
optional, they are not always preserved by aggregator repositories and EuropeanaLocal
technical partners cannot at this moment recommend their use for any purpose outside of the
provider repository itself except to make subsets of the repository contents harvestable for
different aggregators.
9.1.4 Metadata
The metadata section of the record holds the metadata in the actual format. The XML is for
some repositories stored as is, in other as occurrences of an element table.
9.1.5 About (Provenance)

About is an optional and repeatable container to hold data about the metadata part of the
record. The contents of an about container must conform to an XML Schema. About is mainly
used as tracking feature in addition to the identifier.
9.1.6 Flow control

The List verb of the OAI-PMH protocol offers use of resumption tokens to split the transfer of
metadata in smaller chunks. This feature is optional for the provider, but mandatory for the
harvester.
9.1.7 Deleted records

OAI recommends to handle deleted record marked as attribute status=deleted of the header
tag. The Identify verb tells if a repository offers this service or not.
Alternative solutions for this service:
1. The proprietary system marks deleted records as deleted instead of physically removing
them.
2. The proprietary system logs deleted records.
3. An additional system is added on top of the proprietary system to keep track of deletions.
OAI-PMH is described in greater detail in the Implementation Guidelines for the Open
Archives Initiative Protocol for Metadata Harvesting which can be found at the Open
Archive Initiative‟s web site. (Lagoze, Van de Sompel, Nelson, & Warner, 2005).
30/39
online
10 Metadata Extraction and Normalization
Once the repository software has been successfully installed, the next step of the process is to
populate the repository with data extracted from local collection management systems.
Should your local collection management system have an export feature to a well-known
format such as MARC or Dublin Core – or an export to some XML-based format, your best
bet might be to go for this. However, most of the data held by regional and local providers
originates from legacy databases and systems without any such features. For such systems
there are no other ways than writing a custom extraction script to create an ESE formatted
XML representation of the source data.
Metadata extraction
While most of the legacy systems and databases lack any harmonized structure, they are mostly
accessible via an ODBC driver which enables programmers to connect to them and read
content values.
It is too comprehensive a task for this manual to go into the details of the SQL language and it
is assumed that technical personnel performing the metadata extraction is acquainted with this
language and technology. A good references for SQL is to be found at w3schools.com,
(w3schools.com, 2009)
Extraction scripts can be written in any language but most likely one of the web scripting
languages such as PERL, PHP, ASP, Python, VB.NET, C# or Java will be the easiest solution.
After having connected to the database, the extraction script has to do three things:
 Map metadata fields from the source system to ESE
 Normalize any non-standard metadata values
 Create an OAI ListRecords embedded ESE XML file
Metadata Field Mapping
The script must interpret which fields in the source database corresponds t which fields in the
target metadata application profile Europeana Semantic Elements. This may be a task that
requires considerable manual efforts and experimentation in order to understand the practical
consequences of the mapping decisions. In this respect, the ESE metadata elements description
and implementation guidelines will be of great value. See: Appendix 3: Europeana Semantic
Elements v3.1 and Appendix 4: ESE Implementation Guidelines
Metadata Normalization
In addition to reading raw element values from the source database and putting them into
corresponding ESE fields in a text file, the script must also normalize values. The example of
normalization is the adaption of the date format used in the source database to the one
specified by ESE.
Dates can be expressed in a multitude of different ways including the following, which are all
different ways of specifying the 1st of December 2009:
 DD/MM/YYYY – 01/12/2009
 YYYY/DD/MM – 2009/01/12
 YYYY/DD/MM – 2009/12/01
 MM/DD/YY – 12/01/09
31/39
online
In order for a source system to successfully use dates for visualization and retrieval purposes,
it is a strict requirement that date values are normalized.
Another important thing to normalize is the codes used to reference languages. Here
Europeana recommends the use of RFC 4646 (Phillips & Davis, 2006) ISO 639
(Library of Congress, 2008) which defines complete two and three letter
codes for each country and language..
Normalization extends far beyond dates, countries and languages, but once we enter this
territory, it is a requirement that quality assured authority files and thesauri exist from which
we can pick the “normal” value to which other references should be normalized.
This include person names, names of time periods, geographical names and more. All of these
issues will be covered in greater detail in later revisions of this manual. For the time being, it is
required that
No thesauri has been agreed upon for use within Europeana but out of those in existence the
Getty vocabularies (The J. Paul Getty Trust, 2009)are possibly the most promising and
comprehensive resource available and thus a possible starting point. The time and resources
required to perform high quality mapping of content values is however vast and it is beyond
the scope of EuropeanaLocal to embark on this task. The issue will however be covered by
EuropeanaLocal technical workshops.
Writing the XML format
Having determined which values goes where and how they must be split, merged and
normalized to correspond to Europeana Semantic Elements, the next step is to create the
actual XML file which in turn will be ingested into the repository software package installed in
chapter 7 above.
We do this by writing a format which embeds ESE within the OAI-PMH ListRecords format.
The following template may be used and extended with any of the metadata fields listed in
chapter 0 above.
<?xml version="1.0" encoding="UTF-8"?>
<OAI-PMH>
<ListRecords>
<record>
<header>
<identifier>oai:sffKultur.no:SFFkl-100445</identifier>
<dateStamp>2009-02-17</dateStamp>
<setSpec>Kultur</setSpec
</header>
<metadata>
<ese:europeana xmlns:ese="http://europeana.eu/terms"
xmlns:europeana="http://localhost/"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:dcterms="http://purl.org/dc/terms/"
xmlns:dcmitype="http://purl.org/dc/dcmitype/" schemaLocation="
http://europeana.eu/ese.xsd">
<dc:publisher xml:lang=”no”>Fylkesarkivet i Sogn og Fjordane</dc:publisher>
<dc:contributor>Hermund Kleppa</dc:contributor>
<dcterms:issued>2008-11-27</dcterms:issued>
<dcterms:modified>2009-02-17</dcterms:modified>
32/39
online
<dc:type>Text</dc:type>
<dc:identifier><![CDATA[http://www.sffarkiv.no/sffbasar/default.asp?p=resul
t&db=dbatlas_leks&art_id=1280&spraak_id=1&ptype=single]]></dc:identifier>
<dc:source xml:lang=”no”>Fylkesarkivet i Sogn og Fjordane</dc:source>
<dc:language>no</dc:language>
<europeana:isShownAt>http://www.sffarkiv.no/sffbasar/default.asp?p=result&a
mp;db=dbatlas_leks&art_id=1280&spraak_id=1&ptype=single</europe
ana:isShownAt>
<dcterms:spatial type=”text”>Fylke 1400</dcterms:spatial>
<dcterms:spatial type=”text”>Kommune 1430</dcterms:spatial>
<dcterms:spatial type=”sfs-wkt”>SRID=32633;POINT(8361
6822394)</dcterms:spatial>
<dc:rights xml:lang=”no”>Fylkesarkivet i Sogn og Fjordane</dc:rights>
<europeana:unstored xml:lang=”no”>Vegen forbi Matbjørhammaren - smal,
svingete og rasfarleg.</europeana:unstored>
<europeana:type>TEXT</europeana:type>
</ese:europeana>
</metadata>
</record>
33/39
online
11 Testing your Repository
Having completed the previous steps, only one thing remains: to test that your repository is
working. To do so, EuropeanaLocal technical partners provide a test harvester which you can
use to test that your repository is providing valid responses.
The test harvester will connect to your supplied repository URL and issue harvest requests
using the ListRecords verb. If no error messages are returned, the repository is valid and may
be forwarded to Europeana for ingestion into the sand-box environment where the content
provider can see how the exported metadata appears within the Europeana service.
Once satisfied, the repository will be moved to the Europeana prototype and the data store for
the upcoming release of Europeana Version 1.0 scheduled for the spring of 2010.
The test harvesting environment is available at the following URL:
http://utvikling.avinet.no/europeanalocal/harvester2
Additionally, the Open Archives Initiative Repository Explorer (Department of Computer
Science, University of Cape Town, 2006) may be used to verify the functionality of your
provider repository – though this does not allow you to do searches in the harvested data –
merely validate the output of your provider repository.
The harvester is using the PKP Harvester Repository Software Package. Data will be purged at
regular intervals and EuropeanaLocal technical partners offer no particular service level
agreement for this environment which in time will be replaced by the Europeana test
environment once completed.
34/39
online
12 Troubleshooting and Support
Should partners upon attempt to implement the tasks described in this manual run into
problems, EuropeanaLocal offers a number of support options described in greater detail
below.
EuropeanaLocal Support Forum
The online support option for EuropeanaLocal is currently available at

http://www.avinet.no/forum and will by August also be available through the
europeanalocal.eu domain used by the project web site. All partners will receive usernames and
passwords for the forum to the administrative and technical contact e-mail addresses supplied
in the technical questionnaire.
The online support forum consists of five sections out of which two are open for posting from
all users (shown in bold).
 Frequently asked questions: a moderator run section where recurring questions and issues
are highlighted
 Repository installation questions: a community run section where content providers
exchange knowledge among each other supported by the technical partners
 Metadata extraction, normalization and enrichment questions: a community run section
where content providers exchange experience on extraction, normalization and enrichment
issues supported by the technical partners
 Documentation and utilities: a moderator run section where EuropeanaLocal documentation
(like this implementation manual and future revisions) will be made available for download.
 Information from Project Management Board: a non-technical section where content
provider coordinators can give feedback to EuropeanaLocal participants on overall project
progress and related issues.
Other support options
The web based support forum is preferable due to the fact that questions directed there will
benefit not only the originator but also others who experience the same problem. However,
certain issues may be of interest only to individual content providers, in such cases we offer
additional support options.
12.1.1 E-mail support

Where a content provider partner may not wish to expose his/her issues to other partners, e-
mail support is an alternative to using the web support forum. Questions will be answered and
dealt with in a similar fashion as for the support forum, but will not be public. Technical
partners wish to publish questions deemed to be of general interest subject to consent from the
originator, applying anonymization where required to protect privacy and/or security.
Questions may be asked in English, Slovakian, Norwegian, Danish or Swedish to the following
e-mail address:
europeanalocal@avinet.no
35/39
online
E-mails sent to this address will be distributed to the support-team in Avinet and EEA and the
reply will come directly from either of these organizations or be forwarded to the relevant
source if appropriate.
12.1.2 Phone support

Phone support may be agreed upon between content providers and technical partners via e-
mail and will be available from EEA and Avinet subject to limitations due to budget and
resource constraints.
12.1.3 On-site support visits

While deemed a non-strategic support mechanism as it is confined to a very limited number of
beneficiaries, on-site support visits remains an option. However, as an alternative,
EuropeanaLocal has invited technical experts from across Europe to the technical workshops,
thereby ensuring a wide-spread network of OAI-PMH and ESE experts throughout the partner
regions. It is preferable that partners utilize these resources as this has a more long-term
strategic aspect to it than one-of visits from technical partners.
In very particular cases, the project budget gives provisions for a limited number of onsite
support visits. These must be sanctioned by the Project Management Board and will only be
used most or all of the below criteria are satisfied:
1. A partner has un-resolved technical issues
2. The resolution of the issues are time-critical to the project
3. The resolution of the issues cannot be satisfied by any other support options
36/39
online
Bibliography
Clayphan, R., Dekkers, M., Fernie, K., Gradmann, S., Lupovici, C., Meghini, C., et al. (02/2009).
Specification for the Europeana Semantic Elements v3.1. Den Haag: European Digital Library
Foundation (Europeana).
Clayphan, R., Lupovici, C., Sugimoto, G., & Verleyen, J. (05/2009). Metadata Mapping &
Normalisation Guidelines for the Europeana Prototype. Den Haag: European Digital Library
Foundation (Europeana).
Department of Computer Science, University of Cape Town. (2006, 12). Open Archives Initiative -
Repository Explorer - Version - 1.46c. Hentet 04 2009 fra OAI Repository Explorer:
http://re.cs.uct.ac.za/
Lagoze, C., Van de Sompel, H., Nelson, M., & Warner, S. (2005, 05 03). Implementation Guidelines for
the Open Archives Initiative Protocol forMetadata Harvesting v2.0. Hentet 11 04, 2008 fra OAI-PMH
Implementation Guidelines: http://www.openarchives.org/OAI/2.0/guidelines.htm
Library of Congress. (2008, 11 07). Codes for the Representation of Names of Languages. Hentet fra
ISO 639-2 Language Code List - Codes for the representation of names of languages:
http://www.loc.gov/standards/iso639-2/php/code_list.php
Phillips, A., & Davis, M. (2006, 09). RFC 4646 - Tags for Identifying Languages. Hentet fra Network
Working Group, Request for Comments: 4646: http://www.ietf.org/rfc/rfc4646.txt
Powel, A., & Johnston, P. (2003, 04 02). Guidelines for implementing Dublin Core in XML. Retrieved
04 10, 2009, from Dublin Core Metadata Initiative: http://dublincore.org/documents/dc-xml-
guidelines/
Public Knowledge Project. (2009). Open Archives Harvester . Hentet fra Public Knowledge Project:
http://pkp.sfu.ca/?q=harvester
TEL+ Project. (2009). REPOX A Metadata Space Manager. Hentet fra REPOX:
http://repox.ist.utl.pt/index.html
The J. Paul Getty Trust. (2009). Learn about the Getty Vocabularies . Hentet 04 2009 fra Getty
Vocabulary Program (Research at the Getty)::
http://www.getty.edu/research/conducting_research/vocabularies/
w3schools.com. (2009). SQL Tutorial. Hentet 04 05, 2009 fra w3schools.com:

http://www.w3schools.com/sql/default.asp
w3schools.com. (2009). XML Tutorial. Hentet 04 02, 2009 fra w3schools.com:

http://www.w3schools.com/xml/default.asp
37/39
online
Tables and Diagrams

Table 1: Feature comparison matrix for assessed repository software packages............. 15
Table 2: Advantages and drawbacks of repository software packages ............................ 16
Table 3: ESE elements .................................................................................................... 27
38/39
online
Appendices
Appendix 1: Dublin Core Metadata Elements
Appendix 2: OAI Protocol for Metadata Harvesting
Appendix 3: Europeana Semantic Elements v3.1
Appendix 4: ESE Implementation Guidelines
39/39

D4.2 (Modified)

Uploaded by

Copyright:

Available Formats

D4.2 (Modified)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

D4.2 (Modified)

Uploaded by

Copyright:

Available Formats

Technical support documentation and help desk online

Technical support documentation and help

Deliverable number D4.2

Delivery date 21 June 2009

Stein Runar Bergheim, Olav Tuften, Idar

2 DOCUMENT CONVENTIONS ............................................................................................................ 5

3 EUROPEANALOCAL GLOSSARY .................................................................................................... 6

4 OVERVIEW OF THE IMPLEMENTATION PROCESS ................................................................. 12

5 BEFORE YOU BEGIN ....................................................................................................................... 13

12 TROUBLESHOOTING AND SUPPORT .......................................................................................... 35

TABLES AND DIAGRAMS ........................................................................................................................ 38

APPENDIX 1: DUBLIN CORE METADATA ELEMENTS ...................................................................................... 39

Application profile Adaption of a metadata standard for a specific purpose or

Athena A sister-project to EuropeanaLocal seeking to develop

Authentication In the context of Europeana Local, the process of verifying

Authority file An authority file defines a list of known unique entities.

Creative Commons A set of predefined rights expressions which may be used to

Dictionary A list with definitions of words and/or translations of words

Digital collection In the context of Europeana Local, a digital collection is a

Dublin Core Metadata standard, defines a wide range of descriptive

EDLF Abbreviation for the European Digital Library Foundation.

ESE Abbreviation: Europeana Semantic Elements, see

Europeana A web application run and developed by the EDLF. The

Europeana Semantic Elements An application profile of Dublin Core adapted by the

Europeana themes A set of themes proposed by Europeana which should be

GIS Abbreviation for Geographical Information System

Harvesting See metadata harvesting

Harvesting authority A registry contained in a harvesting repository (e.g.

ID Abbreviation for identifier. See identifier, persistent

Identifier A number or string which is used to uniquely identify a

Example: an item might have the ID 23568 in a local

By prefixing a local ID with a unique string it will be unique

IPR Abbreviation for Intellectual Property Rights

Item An Item is one record of metadata describing an abstract or

See: ESE, application profile

Metadata Data describing a resource in terms of intellectual property

Metadata enrichment The process of improving the quality of metadata through

Metadata extraction The process of reading metadata from custom digital

Metadata harvesting The process of reading metadata from a source repository

Metadata mapping The process of mapping the fields of a source metadata

Miniature In the context of Europeana, a miniature is a simplified

MLA Abbreviation for Museums, Libraries and Archives.

If the source vocabularies are not mapped to a common,

OAI Abbreviation for Open Archive Initiative. A body

OAI-ORE Abbreviation for Open Archive Initiative – Object Reuse

and Embedding. A standard for harvesting of rich content

OAI-PMH Abbreviation for Open Archive Initiative – Protocol for

Object An Object is a composite representation of digital content

Ontology A comprehensive set of concepts with relationships

An ontology is instrumental in creating intelligent search

OWL Abbreviation for Ontology Web Language. An XML-based

Persistent identifier An identifier which remains unchanged over time.

RDF Abbreviation for Resource Description Framework. An

Repository Metadata storage software which enables harvesting

download it‟s content in one or more formats. Repository

Schema (database) A database schema is a space in a database where

Schema (metadata) A metadata schema is the same as a metadata format. ESE

Semantic Web A type of web service which allows “intelligent” searches

SKOS XML format for describing a vocabulary, used to facilitate