|
How do users want online databases to work?
By Francisca Hernández, Residencia de Estudiantes,
Madrid and partner
Covax is designing online database systems which
will ultimately be useful to end-users searching the web. Although
our primary focus is to create systems to support interoperability,
resource discovery and access, we conducted a survey in mid 2001
amongst a small but international sample of 'end users' of databases
to find out what they feel they are looking for in terms of functions
and features. We reached the following conclusions, from a questionnaire
survey producing 86 responses from six partner sites in Austria,
Italy, Spain, Sweden and the UK. Those responding were a mixture
of information professionals, academic, clerical staff and others
related to partner institutions in some way. This selection was
necessary to achieve rapid results in the available timescale.
- The educational level, familiarity with the use of computers
and databases, and number of times per week respondents use on-line
databases, is high. More than half of the end users use online
databases more than 5 times a week.
- The main reasons for using and searching databases are academic,
professional or personal research. They are also used for general
information searching.
- Respondents mainly search for full documents or books (58%)
and to a lesser extent for other materials, images and locations.
- The search fields usually used are subject, word in any field,
author and title. The fields related to publishing (date, place,
publisher) are seldom used.
- The operator mainly used for searching is AND (41%) and sometimes
help is needed about its use. It is interesting to note that a
large number of people do not use operators. There are two possible
reasons for this: the users do not know how to how to use operators
or the appropriate operator is not available.
- This questionnaire shows that users wish to customise search
results by changing the display format or result order.
- Ease of search, speed of response and clear results presentation
is what users expect a database to offer. The help function is
less important to them.
- The majority of users (71%) are prepared to wait up to one
minute for the response, but a few are more impatient and they
wouldn't wait for more than 30 seconds, whereas others are ready
to wait 5 minutes if necessary.
- When they need help with database searching, users look for
a help button or search a different way (such as asking a librarian).
- They usually request full display, print or download when they
have found information. Only 24% download information.
- If end users do not retrieve the expected records, they think
this is because they do not exist in the database or they have
not found appropriate search elements. Sometimes they think the
results are too broad and they would need help to effectively
narrow down the search.
- If they do not find an item in a catalogue they try another
search or search another site. A minority of users use the help
button or ask for outside help.
- 69% of the users find it useful or sometimes useful to store
their query for future reference. A third of respondents never
does.
- Two thirds of those polled think it is easy to find what they
are looking for and 25% find it difficult. This result can be
explained by the high level of familiarity with the use of on-line
databases amongst our users.
- They are moderately interested in document reproductions and
perhaps occasionally they would be willing to pay for a reproduction
(61%), depending on the document.
The results of the study will be useful for
our validation and usability assessment phases now underway as well
as informing our exploitation plans.
PARTNER PROFILE: Ente per le Nuove tecnologie
l'Energia e l'Ambiente (ENEA)
By Luciana Bordoni, ENEA, Rome
ENEA is a public corporation operating in the field of research
and innovation for sustainable development, for the purpose of promoting
together the objectives of development, competitiveness, and employment
and that of environmental protection.
It also acts as an agency for public administrations by providing
advanced services in the sectors of energy, the environment, and
technological innovation.
In particular, it:
carries out, develops, and promotes research and innovation, through
the creation of demo systems and pilot projects, with aims and objectives
related to sustainable development, within the framework of the
national research programme and in conformity with the commitments
arising from Italian participation in the European Union and other
international organisations on the subjects of energy, environment,
and technological innovation;
- supports the innovation processes of manufacturing,
in particular of small and medium-sized enterprises, also promoting
demand for research and technology in conformity with the principles
of sustainable development;
- aids the process of technology transfer
and positive experiences in the energy and environmental field
to enterprises, in particular small and medium-sized ones, and
to public administrations within the framework of national policy
and that of the European Union;
- supplies, on request, in the sectors of its competence and within
the framework of programme agreements with the Ministries of Industry,
Commerce, and Handicraft, of Universities, and of Scientific and
Technological Research and the Environment, as well as with other
public administrations, specialised technical and organisational
support to administrations responsible for public activities,
at the national and international level, to regions, and to local
government units.
COVAX metasearch engine and access to native
XML-data bases
By Walter Koch, AIT (Angewandte Informationstechnik Forschungsgesellschaft
mbH), Graz
The COVAX-Project was initially based on proprietary XML-Servers
(TAMINO) holding metadata formatted in XML. These metadata catalogues
from museums, libraries and archives follow different domain specific
standards such as MARC21, AMICO, EAD and TEI. During the project
lifetime other native (the database stores XML records directly)
XML-Servers have become available. One of them was IXIASOFTs "TeXtML"-Server.
This Server is implemented only for the Windows platform, providing
an API (Application Programming Interface) to ASP (Active Server
Page) programs only.
Figure 1 shows the deployment diagram with the different components
included in the COVAX Architecture. To connect the XML-server directly
to the COVAX-component in general the RMI (remote method invocation)-
approach has been chosen. The queries which are distributed by the
COVAX-metasearch-component (the left block in Figure 1) are XPATH-like
(See Figure 2) and are routed via URLs to the WEB-Servers which
are connected to the XML-Databases.
This technique cannot be applied directly to the integration of
the TeXtML-Server. For that reason AIT - Angewandte Informationstechnik
Forschungsgesellschaft mbH - has developed an interface which consists
of two parts: one part (a Windows DLL) supports the TeXtML-API and
is interfaced to an ASP script which resides on a WEB enabled directory
at the WEB-Server's site. By this means a functional integration
of the XML-Server into the COVAX system has been achieved. Another
problem to be solved was the mapping of XPATH- (XQL-) queries to
the standard used by the TeXtML-server. The most frequent query
expressions and search functions have been mapped, so the end user
is not aware of the types of XML-Server to which his query has been
broadcast. The query in Figure 2 (the substring "Aus"
in the title field) had to be transformed to the form as outlined
in Figure 3 (<anystr/> means "wildcard"-sign):
The semantic mapping from the DublinCore (DC) data element: dc-title
to the AMICO-dtd relevant element: OTN__object_title_name is done
within the COVAX-system on the user interface side. Another option
for this mapping is to do it at the index-level of the XML-Server
(see Figure 4). "Key NAME" corresponds to "index
NAME".
A hit delivered by the XML-Server (XML-formatted) is represented
in Figure 5 (an example from the Basel Poster Collection).
The integration of the TeXtML-Server into the COVAX-system has
- besides the fact that the potential user of the COVAX-System has
more than one option for an XML-Server - another benefit. Many utilities
are available supporting the TeXtML-Server, and the powerful API
enables fast development of (ASP-based) applications. For example,
based on a TeXtML-Demo-Package, AIT has developed a simple (meta-)
data management and search facility for single data bases. Using
this tool (Figure 6 and 7), the COVAX-user (who runs the data base)
can manage his/her own data base, which is also accessed via the
the metasearch engine without any additional effort.
AIT has in addition developed an installation tool which allows
the setup and management of different data bases which can be searched
in parallel via the COVAX Metasearch engine. This possibility is
already used within the Culture-2000 project VECTOR. In this project
different archives all over Europe are maintaining their different
catalogues separately. The "Virtual Catalogue" again is
achieved by the use of the COVAX-System.
TEXTML at Blekinge Tekniska Högskola
By Peter Linde, Blekinge Tekniska Högskola, Karlskrona
We decided to use IXIASOFT's TEXTML-server, version 1.6 (http://www.ixiasoft.com)
this spring after receiving favourable reports about the software's
performance from our COVAX partner AIT in Graz.
The main reason for installing and running the TEXTML-server package
at BTH, except of course for doing what we wanted it to do, was
the price. Unfortunately the price went up from 3.000 USD to 10.000
USD just before we decided to purchase, but even so it compares
very favourably with TAMINO if you only look at the price.
It also has proven itself quite easy to install and start running.
The only negative remark we make is regarding the administrative
user interface, which is not very user friendly. For example, it
is quite complicated to manipulate records in the different document
bases - to show, delete, transfer etc.
All communication with the database is done via different indexes.
This gives a lot of possibilities to do searches on xml-records
at even the lowest attribute level, which suits us fine since the
main bulk of our records are full text documents, mostly fiction,
tagged for research purposes in TEI.
But indexing also creates some problems that we have encountered
during the COVAX project. To retrieve records via the Metasearch
engine all the partners uing TEXTML must have corresponding indexes
that also match all the accesspoints for the TAMINO-partners. This
has been shown to be crucial and has raised problems with the early
prototypes of the COVAX system.
Our profile in the project has been to create and archive fulltext
records. During our tests it has been proven quite timeconsuming
to shuffle these big files back and forth from the partner repository
to the main COVAX search site. This has resulted in long wait times.
To solve this problem we have developed a search-and-display interface
for our TEXTML-database (for display we used Sebastian Rahtz's XSL
stylesheets - http://www.tei-c.org/Stylesheets/
- with minor modifications). Thus we can produce HTML versions of
our fulltext documents.
Being able to do this, plus being able to do searching within the
document bases will make it possible to instruct the COVAX search
engine only to retrieve document metadata, not the full document.
In the metadata there will be a pointer to the full document at
our site. So when the user searches COVAX and finds any of our full
text documents, the user will find a link to the full text document
at our site. If he follows the link it will take him to the full
document, which he now can investigate with further searches directly
from our own web-interface based on the elements used for the full-text
documents.
We are currently running the TEXTML software on a Dell PowerEdge
1300, under Windows 2000. As soon as the COVAX project allows us
we intend to upgrade to version 2.0 which was released in September
2001.
EVENTS
"Safeguarding our culture", Society of Archivists' Annual
Conference, Aberystwyth, UK, 4-7 September 2001
The UK's Society of Archivists kindly invited Robin Yeates to present
a paper about Covax to their main annual meeting, held in the normally
very pleasant but sometimes wet, Welsh, holiday, university and
seaside town of Aberystwyth.
The paper was entitled 'Using XML to make cultural resources accessible
- the COVAX project'. Its theme of integrating library, archive
and museum resources obviously struck a chord with many of the audience.
However, the short time to explain the technical approach of the
XML-based project meant that some people may have had difficulty
understanding it all!
As a technology project primarily concerned with developing appropriate
'middleware' for networks, it is becoming increasingly difficult
to introduce this kind of work to outsiders, but those trying to
achieve similar goals were extremely interested to receive further
information after the event.
If you are using XML technologies to exchange and make available
cultural resource data, contact us to find out more, once you have
read this newsletter of course...
The Society of Archivists has a website at
http://www.archives.org.uk/
COVAX metasearch engine and access to native XML-data bases
Workshop on Artificial Intelligence for Cultural Heritage and Digital
Libraries
AI*IA 2001 Conference, Bari, Italy, 25 September 2001
This workshop represents an opportunity to analyse the contribution
that the Artificial Intelligence and the Web could bring in this
contest and, also, to directly compare ideas and results coming
from academic and research corporations and from companies/bureau
that are involved in developing new technological solutions in the
field of cultural heritage and of digital libraries. The goal is
to create a community of institutions that could deal with this
new challenge. In this context Luciana Bordoni presented in the
paper entitled "The Covax project" the main aims of Covax.
(http://www.di.uniba.it/~aiia/ws2/ws2_en/)
Online Information 2001 - London, UK
4-6 December, 2001
http://www.online-information.co.uk/
LITC is an exhibitor at this major event and Robin Yeates will be
present to discuss Covax.
An overview article about Covax by Robin Yeates will appear in a
special VINE issue to be published for this event and available
at the stand.
Contacts and information
COVAX-IST-1999-11820
A website has been established in Catalan, English, German, Italian,
Spanish, and Swedish at http://www.covax.org,
where you can find other issues of Covax News and other information
about the project.
Project coordinator
Carlos Wert
Residencia de Estudiantes
Pinar, 23
28006 Madrid
Spain.
Email: ile@interlink.es
URL: http://www.covax.org
Rights and disclaimer
COVAX is part funded by the European Commission.
The authors are solely responsible for the content of this newsletter.
The site does not represent the opinion of the Community and the
Community is not responsible for any use that might be made of data
appearing herein.
Reproduction of the newsletter is authorised, except for commercial
purposes, provided the source is acknowledged.
Some of the documents and multimedia sequences in this newsletter
and on the COVAX website might contain references, or pointers,
to information maintained by other organisations. Please note that
we do not control and cannot guarantee the relevance, timeliness
or accuracy of these outside materials.
© COVAX 2001
Dissemination
Robin Yeates
Associate Director
LITC, South Bank University
Postal address: 103 Borough Rd, London SE1 0AA United Kingdom
Tel: +44 (020) 7815 6924
Fax: +44 (020) 7815 7050
email: yeatesrb@sbu.ac.uk
URL: http://www.sbu.ac.uk/litc
More info
If you have any questions regarding COVAX or want to receive more
information about its activities and news please click
here
|
