Docstoc

external review Executive summary

Document Sample
external review Executive summary Powered By Docstoc
					EIARD Infosys Phase II

Assessment report




Chris Addison

February 2001
EIARD Infosys phase II Assessment report                                 C Addison       February 2001




Contents
INTRODUCTION                                                                                             3
  METHODOLOGY ............................................................................................ 4
  INFORMATION SYSTEMS .................................................................................. 4
1 GOVERNANCE AND MANAGEMENT                                                                              5
  1.1 REPRESENTATION OF EIARD MEMBERS .......................................................... 7
  1.2 THE DEVELOPMENT PROCESS AND EUROPEAN COHERENCE ..................................... 9
  1.3 STANDARDS .......................................................................................... 9
  1.4 SUSTAINABILITY ................................................................................... 11
2 IMPACT                                                                                              12
  2.1 PROVIDING INFORMATION TO POLICY MAKERS IN EIARD TO HELP DECISION MAKING. . 12
  2.2 EUROPEAN INPUT TO INTERGOVERNMENTAL AND MULTILATERAL INITIATIVES. ............ 12
  2.3 PRESENTATION OF DATA SOURCES .............................................................. 13
3 USER PROFILES NUMBERS AND TARGET.                                                                    14
  3.1 NUMBERS OF USERS .............................................................................. 14
  3.2 TYPES OF REQUEST ................................................................................ 15
4) FUNCTIONAL STRUCTURE AS A DECENTRALISED SYSTEM                                                      17

5 FUTURE PROSPECTS FOR INFOSYS                                                                         17
  5.1 PROSPECTS FOR EIARD-INFOSYS .............................................................. 17
  5.2 OPTIONS FOR THE INFOSYS STRATEGY ......................................................... 18
  5.3 DOES ARD IN EUROPE REQUIRE A SPECIALISED INFORMATION SYSTEM ................... 18
  5.4 THE INFOSYS NICHE .............................................................................. 19
6 TECHNOLOGY                                                                                           20
  6.1 EXTENSIBLE MARKUP LANGUAGE (XML) AND INFOSYS ..................................... 20
  6.2 MULTISEARCH POSSIBILITIES .................................................................... 21
7 SUSTAINABILITY                                                                             22
  7.1 DEVOLUTION OF INFOSYS FUNCTIONS ......................................................... 22
  7.2 MEMBERSHIP MODELS AND FEES ................................................................ 23
  7.3 ESTABLISHING A DONOR GROUP ................................................................ 23
  7.4 NETWORK OF EXCELLENCE ....................................................................... 24
GLOSSARY                                                                                               25
  ANNEX 1
  STRENGTHS WEAKNESS OPPORTUNITIES AND THREATS ANALYSIS PREPARED FOR CORE
  GROUP.
  ANNEX 2
  PRESENTATION TO CORE GROUP.




                                                                                                            2
EIARD Infosys phase II Assessment report                  C Addison   February 2001




Introduction

EIARD-InfoSys is an Internet service featuring European Agricultural Research for
Development (ARD) resources. This service is the collective activity of National
nodes from European member states and in addition Norway, Switzerland working
together with associated information networks.
The project has the following joint mission:
"EIARD-InfoSys aims at increasing the transparency of and access to European
web resources in the areas of agriculture, environment, forestry, fisheries, socio-
economics, rural transformation and many others, devoted towards development.
It also aims at creating an information and communications platform, as a service
for a multitude of institutions and parties all over Europe involved in scientific
development cooperation."


This report covers a review of the latest phase of the project in the following
seven areas.


       1) Governance and Management
       2) Impact contributing to decision making, input to intergovernmental and
       multilateral initiatives and presenting data in accessible form.
       3) User profiles numbers and target.
       4) Functional structure as a decentralised system.
       5) Future prospects for InfoSys in light of other services.
       6) Technology in particular use of XML
       7) Examine sustainability


The European Initiative for Agricultural Research for Development (EIARD)
commissioned this study. This group serves as a mechanism that enables better
coherence between relevant policies in ARD within the European Member states
and in addition Norway and Switzerland.
The InfoSys activity comprises the following services: The central InfoSys service,
http://www.eiard-infosys.org, National node websites, support for European and
National Fora web services, the EIARD and European Forum for ARD websites.
The resources featured in the main service cover
A) News, Events, Discussion fora, training courses, experts, Funding
opportunities, Institutions, Projects, Networks, Information Systems,
and
B) Signposting to sites for additional Projects, Organisations, Publications, news
services and Job Vacancies.
These are not just web pages but manually entered indexes of resources. Entered
and updated by National nodes, or by general users and reviewed the resources
cover all EU member states, Norway and Switzerland.
The resources are entered directly in InfoSys supported systems in the case of
group A and signposted to separate systems in group B.




                                                                                      3
EIARD Infosys phase II Assessment report                  C Addison    February 2001


The user can search those in A directly from the InfoSys system but does not
presently have a way of directly searching the contents of all the information
services indexed in group B.



Methodology
The terms of reference and methodology were discussed with the EIARD secretary
and the coordination unit together with the identification of stakeholders who
would be approached for interview.
The evaluation began with a review of the site and documents. The general
description of the site and operation of the service was then elaborated with the
EIARD secretary and a site visit to the Coordination unit. As the multisearch
project was a substantial part of the review a site visit to IAO the Italian National
node was required together with a visit to the Global Forum for Agricultural
Research.
To clarify the application of the terms of reference the EIARD working group
members were approached for their perceptions of the project. Having established
the framework of the project and current use all the National nodes and
associated networks were approached with the questions most relevant for their
activities. This avoided redundancy in the questions posed and ensured more
focussed feedback.
Substantial face to face interviews were summarised as numbered points and sent
by email to the participants for validation.
This differed from a normal evaluation in being a review of the project against
criteria not involved in the existing project. It was therefore a review of the
potential of the project to contribute to these terms of reference.
Without a dedicated workshop for the process individual views had to be
assembled in a framework on which others could comment this was done as a
SWOT analysis against the terms of reference which was presented at a core
group meeting in Bonn (18-19 February 2002) to elicit feedback. The paper was
seen at the meeting as a fair reflection of the potential of the project and current
situation. Detailed analysis of the management, statistics, added value of the
information service and the funding possibilities were handled separately.
Additional information for the future options are drawn from personal experience
of similar Internet services and library and web research.
Evaluating the project activities of the past two years it should be clear that few
of these terms of reference were addressed in the project. This paper therefore
represents a review of past and potential activities in the current environment
surrounding InfoSys, in terms of resources, content, technology, partners, actors
and players.


Information systems
The approach of using a web service as a system of information support to a
specific European Research topic such as Agricultural Research for Development is
throughout Europe networks. The subject of Development Cooperation differs
from most activities in requiring interaction with global users even more essential.
A number of current European commission supported activities play a similar role.
These vary from individual programmes such as Asia Urbs which offers a
dedicated partner locator service on the internet together with details on
individual EU projects to whole networks supporting web resources in one subject
area, for example EUFORIC providing Europe's Forum for International
Cooperation.



                                                                                        4
EIARD Infosys phase II Assessment report                 C Addison    February 2001


InfoSys differs in engaging a variety of players in the provision of information and
not leaving it completely to active end users to submit information or inputting
information centrally. In this respect the information base can be more
comprehensive than services relying on either a totally centralised or totally end
user basis.
The complication of such a service is that it is not based on one central database
but instead attempts to provide access to a number of databases.
This brings the disadvantage that the user has more work to do in retrieving
information, but the advantage that there is more information available. The
quality of the information may vary but once a user has identified the relevant
dataset it may be in a more appropriate form.
The system has relied on the fact that there are a variety of approaches to
managing the data. Information that is used within organisations such as mailing
lists will need to be held and maintained onsite. Information that is used for
promotion such as training opportunities may be held externally on one database.
Therefore there are a number of significant project database collections, German,
Italian and Belgian collections. together with 600 European records from Cordis.
InfoSys has built a substantial index of most of the relevant data sets in Europe.
The user now needs an easier navigation and tools to retrieve data from these
systems.
There are two solutions:
a) to collect all the data in one database
b) to provide a way for the user to search all the databases.
InfoSys has chosen to use the two models depending on the type of data and
partner.
For National nodes it is transferring metadata on links and organisations to the
Agriscout database or asking for direct input.
For sectoral information it is also pointing to other Organisation databases, e.g.
ETFRN.
For project information it asks organisations without existing project databases to
input to the Agriscout database. Those who already have project databases online
are invited to join the multisearch project which should provide a prototype
service by the end of the project.
InfoSys has focussed on being an indexed site providing a window on existing
web resources. To become a service it would benefit from orientation towards
specific products supporting the EIARD ideals for example a partnership window
to identify partners in ARD. An email alert service to raise awareness of events
and news as well as content.




1 Governance and Management

The EIARD InfoSys project is an Accompanying Measure of the Fifth Framework
Programme of the European Commission. It does not have a legal status in its
own right, and although linked with is not formally a project of EIARD. However,
the EIARD Working Group closely monitors the project on behalf of EIARD,
particularly at the strategic and policy level. EIARD members also facilitate the
operations of the national nodes in their respective countries. The continuing
support of EIARD is therefore essential to InfoSys.



                                                                                      5
EIARD Infosys phase II Assessment report                 C Addison    February 2001




The Steering committee is responsible for guiding and overseeing the overall
project. It has specific responsibilities as an advisory board;
to support the launching of InfosysII; to improve the synergy between EIARD and
InfoSys; to promote support to National nodes; to identify related information
activities and to develop partnerships


The EIARD-InfoSys Core group consists of representatives of the National nodes
(which have been nominated by EIARD) and members of European Institutions
and networks. This group is responsible for steering the project through annual
meetings in a workshop environment. The Core Group comprising the European
Commission project manager, National nodes, the Coordinating Unit and the
EIARD Secretary. Associated partners such as the networks are invited to the
meetings.




The system itself is then managed by a small coordination unit supporting the
technical developments, communications functions and the contacts and co-
operation between National nodes and with external partners. The Coordination
unit is also responsible for ensuring regular interaction between the EIARD-
InfoSys Core Group by organising annual meetings and providing information and
feedback. The Coordination Unit raises issues with the Steering Group for
decisions proposed to the Core Group.


The activities within the project contract with the commission have been
completed as agreed. The coordination unit has successfully managed the
development of these technical elements. The individual activities were reviewed
with the Coordinating unit and a project schedule record1 produced showing the
completion of tasks and remaining scheduled tasks. The only area of concern on
the basis of the contract was the abandoning of different language interfaces.
The present project did not directly reflect the needs of the EIARD working group
and there were few indicators to ensure that these requirements would be met.
Even the basic assumption of the EIARD Working group and the European
Commission that the service would provide a comprehensive list of Organisations
active in ARD in Europe has not been met. This is not a failure of the project
directly but of the differing perceptions of the various players and the lack of
indicators.
Any future project must serve the needs of EIARD more closely and contain at
least basic targets for content development and user satisfaction.
The coordination unit is very reliant on technical support for the Agriscout
database on ZADI, to the extent that reports on content from the database were
not available for analysis. This is being addressed by the project by the migration
of Agriscout to an SQL platform which can be managed directly by the
coordinating unit. Similar problems exist with the production of web statistics. For
most of 2001 web statistics were not available and although running now do not
contain DNS lookup information meaning that there can be no analysis on where
visitors to the site have come from.
This type of information is valuable for recording changes in the use of the site as
well as an instant view on the data so it is important that a hosting contract
defines the services that InfoSys is paying for from ZADI and ensures they are

1
    Project schedule annex



                                                                                      6
EIARD Infosys phase II Assessment report                  C Addison    February 2001


delivered. It was also noted that during the time of the review the EIARD website
was unavailable for at least two days and the EIARD-InfoSys site database and
site were unavailable for at least one day. Statistics on total uptime as well as the
production logs of the web statistics should be supplied to the Steering group at
least once a year.


1.1 Representation of EIARD members

National Nodes
National participation has varied. (Some are funded by Governments, some are
not) There are two types of participation, a) Nodes who do not have existing
systems for handling National ARD information where data has been directly
entered into EIARD InfoSys, all nodes have some entries made in this way. b)
Those who have national databases already and have loaded this into the central
system, IAO in Italy and Infoagrar in Switzerland work this way. For IAO this
batch loading of materials is in place and operated at present on an as needed
basis. In the case of Infoagrar this is still a double entry process. This question of
data transfer needs to be addressed to ensure regular data input where existing
databases hold the relevant information. This is likely to use the technology of the
multihost system. Number of entries in the dataset vary widely, but this is to be
expected when realising that Ireland for example may have 20 individuals
involved in Agricultural Research for Development, whilst Germany has over
4000. The National nodes have clear Terms of reference. These are rather over
ambitious for most nodes so are adopted selectively, and might be worth
reviewing to a more realistic level.


The two countries who are not representative are France and the United Kingdom.
Relevant information providers have been identified for this data but National
resourcing for this data input is not currently available. This is a two stage
process, arranging the initial data take on and making arrangements for
maintaining the data.


National input to the EIARD InfoSys
The number of entries by each country was taken from the Agriscout system. This
shows the number of entries of a combination of Projects, organisations and
metadata for each country. These are not representative of the true resources
available, in the case of France and the UK. For the other countries there needs
to be acceptance by National representatives as to how accurate the view
currently is. Having compared with previous directories of research organisations
working in Development the InfoSys entries are more accurate than the Printed
World directories where there are no more than 20 European entries for ARD.




                                                                                       7
EIARD Infosys phase II Assessment report                               C Addison   February 2001


Numbers of entries in Metadata system - AgriScout




      GERMANY


           UK


   SWITZERLAND


         ITALY


   NETHERLANDS


       AUSTRIA


       BELGIUM


       FRANCE

         SPAIN


        EUROPE


       IRELAND


      DENMARK


       FINLAND


      NORWAY


       SWEDEN


     PORTUGAL


        GREECE


                 0   100   200   300   400   500   600   700   800   900   1000




Associated networks
There are a number of related information networks some falling within the scope
of InfoSys and others overlapping. The European Tropical Forest Research
Network (ETFRN) and the European Integrated Pest management Working Group
(IPMEurope), are examples of networks working in ARD. Their information is
featured in InfoSys, but has become more integrated into the service. ETFRN for
example has a substantial database of tropical forest research organisations
which are featured directly from the Organisation pages of InfoSys.
Other associated Information networks contain relevant materials. The WISARD
system for example contains a number of project and organisation records.
These groups have a more unclear status in the decision making process. The
latest core group meeting with associated partners raised the concern for the
need for a Memorandum of Understanding, or at the minimum a letter on intent
between InfoSys and these networks.
These partners are not only critical to ensure that InfoSys is comprehensive in
content, but also to ensure that InfoSys is promoted to all interested parties.


Future
As InfoSys grows the governance structure will need to develop, concerns have
been raised as to how the membership of the Steering group is determined in the
future. The question will be whether associated networks are members of this
core group or associate partners.




                                                                                                   8
EIARD Infosys phase II Assessment report                  C Addison   February 2001


Memoranda of Understanding or letters of intent with the networks need to be
developed to avoid confusion and be clear on ownership of data, whether owned
by the project or by the contributor.
The network of Information managers dealing with ARD information resources
also promote ARD through their own services.
InfoSys is an overall view of ARD services. The network includes a variety of
expertise in delivering this information to different audiences which is available as
a resource to developing country regions, this needs to be harnessed by InfoSys.


1.2 The development process and European coherence
InfoSys has provided a platform to present and promote existing services to assist
the development process. InfoSys signposts over 3500 sources of information and
expertise in Europe with a significant usage on the web.
This activity has raised awareness of European ARD resources and better contacts
and linkages of in-country activities through establishing National Nodes.
The present system however does not serve the aims of EIARD as well as it
might. InfoSys still requires significant effort on the part of the user to identify
resources. For example selecting organisations in “Who? What? Where?” does not
then allow selection by subject.
Through further linkages with associated partners such as Euforic and GDNET
there is an opportunity to raise the role of ARD issues in the development process
as a whole.
InfoSys needs to establish where it has a clear value added role. It should be
linked from the CORDIS partner service page for example as a support
mechanism for finding partners for joint projects and work.
Networking between European organisations can be enhanced through the service
directly but also by indirect actions such as the delivery of the news and events
service.


1.3 Standards

A series of data quality and content standards have been developed within the
project for the presentation of information. These have not been promoted
externally and whilst drawing on existing standards with a workshop 2, there has
not been a continuing dialogue with the broader standards discussions.
This is important more recently with the development by the CGIAR centres of a
metadata standard. The Agricultural Metadata Element Set (AgMES) namespace
defines only the elements and schemes that are additional to the Agstandards
definitions.3 The idea is that the elements used in various applications can be
registered in the AgMES domain. The idea behind this is to guarantee a degree of
compatibility between different application profiles.
This would have implications for the possibilities of the Multiple search engine,
but also from the point of sharing practical experience with the groups developing
standards.
A similar situation exists for Project information. The Development Gateway,
originally conceived by the World Bank, has developed a joint database of
projects from donor databases called AIDA. The AIDA project is attempting to
combine project data sets in one search. This project is looking to work with some
2
    AgriXML http://www.agrixml.org
3
    Agmes project http://www.fao.org/agris/agmes



                                                                                      9
EIARD Infosys phase II Assessment report                        C Addison   February 2001


partners of EIARD InfoSys, for example WISARD. The model for providing this
search may be different and not suit InfoSys requirements, but a dialogue with
the group may provide some resources for the development of an ARD window on
existing data.
As a clear step forward the existing standards and guidelines for data should have
a dedicated section in the documentation section, or ideally an individual page,
the URL of which can be linked from the main initiatives. This would mean the
project should be linked from the IDML initiative web site, and the AgMES or
agristandards page.
Otherwise there is a danger that competing standards will be promoted in the
Agricultural and development communities, particularly in the development of the
Regional Agricultural Information Services of other regions.
The current topics list has been developed over the course of the project and is
based more closely on the OECD macrothesaurus. Links with the topic lists used
in other communities become more important as other networks are featured.
ETFRN and EUFORIC for example have their own topic lists. A simple action would
be to submit the existing lists to the FAO ontology server project4 to ensure any
developments take account of the needs of the ARD community. In linking with
an existing initiative it would also be easier to provide multilingual retrieval. The
OECD macrothesaurus provides language equivalents for many of the terms so
would allow possibilities for different language retrieval.
Currently the topic list is skewed in the use that has been made for indexing. As
the Warwick University Study of 19965 suggested, with an indexing service for
Agricultural Research for Development it is important that the topic list can adjust
to the collection. For this reason the breakdown of content needs to be reviewed
to ensure the user can access all the materials as simply as possible.
The project has come to the conclusion at present that language versions of the
site will not be developed. This issue could be handled by the National Nodes in
those countries needing language support for retrieval.




Subject breakdown in Agriscout service

The following diagram shows the breakdown of indexing by subject in the
Metadatabase service Agriscout. This illustrates the change in balance towards
Natural Resources and Socio Economics but also highlights the uneven
distribution of top level terms for example having horticulture as a top level term.




4
    Ontology service Project AOS http://www.fao.org/agris/aos
5
    EIARD-Infosys (Warwick Study; Jan/1996)



                                                                                        10
EIARD Infosys phase II Assessment report                C Addison         February 2001




                                    Veterinary
                                       2%


                                                            Agriculture
               Soc Econ                                       25%
                 25%




                                                                                   Fisheries
                                                                                      4%
                                                                                Forestry
                                                                                    7%
                          Nat Res                                         Horticulture
                           28%                             Nutrition          3%
                                                             6%




Chart showing the percentage of the 3500+ entries in the Agriscout
system with the subject headings they are indexed by.


1.4 Sustainability
The project has secured funding for data input through National nodes, and
associated projects, but there is no direct contribution from partners for the
central structure. This needs to be seen in the context of all the work undertaken
in all the systems featured in the system.
Resourcing should be seen in the context of all the funding going into indexing
functions in the various data sets signposted in the system. The development of
joint programming activities, such as the financial support from Italy for the
development of nodexml in combination with the coordination unit work on the
gateway programming for combining results, points to how further developments
can be resourced. The switch towards this type of joint project and the move
towards featuring existing systems such as the Infoagrar Training database rather
than creating a new central database all point to possibilities for sustainability.
Other models were introduced to the project partners but there is no real parallel
with other network or membership models.


In examining the sustainability the key maintenance functions are for the web
service as a whole, but also the central databases. This maintenance has been
complicated by the split across two platforms involving two database systems and
two servers. The split was necessary as the later databases ran on new software
SQL server, whilst the main database remained on BASIS software. In
consolidating this server to one database and one server as is planned the costs
of maintenance and hosting can be reduced. Where existing databases can be
adopted the sustainability of the approach can be improved, this has to balanced
against the adoption of common keywords and search technologies so that the
information on one topic can be clearly displayed through one view to the user. If
every system is run by an associated network of InfoSys, unless there are links
with their indexing systems views across the system will be impossible.


In approaching the present partners there were possibilities for charging for
particular services or views on the data. For example partner organisations might
be interested in paying an annual charge to be able to provide a language or



                                                                                           11
EIARD Infosys phase II Assessment report                 C Addison    February 2001


subject view on the data. See also section 7 for further comments on the future
situation.


2 Impact

The impact of InfoSys was examined in three areas:
a) Providing information to Policy makers in EIARD to help decision making
b) Complementing and adding value to the global picture in providing European
input to other intergovernmental and multilateral initiatives
c) Presenting data from a combination of national sources in an accessible and
useful way.



2.1 Providing information to policy makers in EIARD to help decision
making.
The present system provides links to source data and so can be used as a first
step for identifying sources of information in ARD. Ten percent of the current
users describe their work as being related to Policy, see section 3. The service
features policy relevant information, but provides policy papers and individual
documents through the EIARD intranet to EIARD working group members. This is
not necessarily used directly by many working group members but provides an
effective repository for the group and in particular for use by the working group
secretary.
The EIARD working group is supported by a website as well as the intranet
supported by the coordination unit. This provides a document repository and
discussion forum for the EIARD working group and a platform for promoting the
EIARD group.
The news service forms an important feed for policy makers and would reach
more people as an email alert rather than relying on visitors to the site. This
would also serve to promote the site.


2.2 European input to intergovernmental and multilateral initiatives.
The present system has a high recognition from the Global Forum for Agricultural
Research. They consider that the InfoSys model is valuable for the networking of
National databases in a regional system for two main reasons:
a) As an organisational model national to region.
b) In providing access to information for a variety of users. InfoSys does not just
offer technical information but a more heterogenous offer.
GFAR also considers InfoSys is not just an Information service but a platform
where Information management issues can be discussed, for example; how to
assess impact on policy, which indicators can be used, how to engage users and
how to evolve to a Knowledge system. Key to this being the principle of users
leaving information behind them whilst using the system.
The input of the InfoSys group to the discussions on creating a global ARD
knowledge pool have been appreciated by GFAR.
The discussions with InfoSys project members have lead to the development of
ideas on the creation of a Knowledge pool. This pool would comprise not just
data, but other inputs from users which they contribute when using the service.




                                                                                  12
EIARD Infosys phase II Assessment report                 C Addison    February 2001


GFAR feels that there are three areas which can be addressed by knowledge pools
in the ARD sector; policy , NARS research management and Technological issues
(such as water management, soil conservation, collective area management)
Each of these pools requires a different construction. The most active area for
GFAR is in Natural Resources Management which is being progressed with NGOs
accumulating knowledge.
The Information tools developed by InfoSys so far are useful for: Policy makers in
general, make less contribution to the NARS dimension at present. The tools also
support technical groups at the topics level where capacity and experience
accumulate through the linkage of resources.
GFAR is assisting in generating Inter regional interaction between Europe and four
regions Asia/Pacific, West-Asia and North Africa, Sub-Saharan Africa, Latin
America and the Caribbean. In particular the emphasis is on engaging the
Information managers and develop coordination.
The decentralised approach adopted by InfoSys is increasingly accepted, counter
arguments are usually based on standards and quality concerns. Other systems
such as WAICENT are now examining the decentralised approach.
The InfoSys coordination unit has played an active role in the Regional
Agricultural Information Service meetings and should continue to do so.
The recent technical developments nodexml are available as Open source
software and available to other projects internationally, for example the FAO
could take advantage of this code in their Multiple search program. More links
with these types of initiative, also AIDA would benefit both parties. The
combination of approaches would improve practical applications, see earlier under
standards.
The engagement with these other projects would also improve the visibility of
European funded projects in ARD.
Contributing to International initiatives may not just be achieved directly by
InfoSys, it is also important to promote European resources in the global search
engines and directories, for this reason some sharing of experience in promoting
ARD resources on the internet would be useful, either through the EIARD InfoSys
newsletter or the documents section of InfoSys.


2.3 Presentation of data sources
EIARD InfoSys relies on frame technology to be able to provide a navigation
system to encompass a whole range of data services.
The current system is a vast improvement over the previous interface and
provides a logical structure to navigate through the site. The initial page is
however very complex and some features in the Agriscout part of the site do not
work with all browsers. This will be addressed with the migration to the new
database platform.
The problemmatic areas of the site run on Basis software, it is hoped that these
limitations will be dealt with in the move this year to an SQL platform.
The current search facilities do not allow a search of organisations by a particular
subject. Within the Agriscout search the user can select to just view organisations
but not then select a subject. There are several search limitations with the narrow
your search option only providing country or geographical zone options to limit
the search. The move to a common platform for all the data will ensure that all
the relevant information can be quickly viewed by the user and in addition more
advanced search features can be introduced.




                                                                                   13
EIARD Infosys phase II Assessment report                              C Addison      February 2001


The formulation of simple URLs to allow linking from other sites so that the
InfoSys identity remains would be useful. This would also aid promotion of certain
services. for example www.eiard-infosys.org/jobs or /news.
The incomplete input from National nodes should be addressed by seeking
additional input from Associated networks, duplication could be avoided through
the validation software. The input of National data can only be enforced if the
National Forum if existing or EIARD working group member can confirm the
entries are an accurate reflection of the National situation, this information can
then be used for policy decisions and fairer discovery of partners.


3 User profiles numbers and target.

3.1 Numbers of users
The User base has increased significantly over the course of the recent phase,
with hits increasing 700% between March 2000 and November 2001.
Extrapolating from this level of hits this would suggest 100+ policy development
user sessions per month, 500+ Researcher sessions, 260 Information brokers and
182 Education sessions, just from those addresses which can be identified as
Europe. The present statistics (end of 2001/2002) do not show a breakdown of
visitors by country, the figures below relate to the number of user sessions in one
month in the previous year.


Number of Users to total EIARD InfoSys site broken down by Country
(note this does not include visitors from the countries who are registered with .com, .org .net and
other addresses without location.)




This level of use will have changed and a request has been made for future
statistics to include DNS lookup to identify the visitors by country.




                                                                                                      14
EIARD Infosys phase II Assessment report                       C Addison   February 2001


3.2 Types of request
It was also not possible to analyse the requests made of the databases showing in
detail all the requests made of the system. This could be analysed from raw logs
to identify the most frequently asked questions which could then form part of the
help system for the service and help in identifying future changes in design.
The types of requests handled in a “Frequently Asked Questions” page could
include the answers provided to the EIARD working group questions posed last
year.


The site has been extensively promoted with the search engines since the User
Study by InfoAgrar in 2001 and the referrals from search engines indicates this
has been successful, however the system could be registered in more agricultural
directories, particularly at an international level. This can be reviewed against the
number of links made directly to the service.


User session roles
Forms completed online from the InfoSys site show the following balance of user
roles. The majority of use is by Information intermediaries identifying sources
which may then be used as a basis for the production of review materials or
research outputs. This tallies with the findings of the User survey showing 20.5%
research, 15.4% education, 14.3% providing information.


Percentage breakdown of user roles
as determined from input forms on the website


                          8%
                                           3%            13%

               14%
                                                                              18%




              8%
                                                                    8%
                         14%
                                                   14%



                Consultant                      Research
                Managing Research               Coordinating
                Information Delivery            Information Dissemination
                Training Education              Policy development
                Marketing


User Study and User Groups
In the next phase, the formation of example user groups even if only a small
sample, for example 15 - 20. would be essential for regular feedback, the results
from the user study show how hard it was to identify a larger group. This would at
least cover three groups, policy, research and education. This could be organised
in conjunction with National nodes.



                                                                                       15
EIARD Infosys phase II Assessment report                 C Addison         February 2001


This would have been a useful mechanism to examine the user satisfaction with
the changes made in reaction to the 2001 User study. There were six
requirements identified from the user study.
a) To promote the system and strengthen National Nodes
The system was promoted in the main search engines, there was a significant
increase in visitors as reported above. Systems to support the National node such
as metadata entry were improved.
b) Provide simple and fast search tools, enhance current structure
The InfoSys service has improved in terms of facilities and services. It now
provides a substantial platform for the development of content. The home page
has become more complex with the addition of news and events but the general
navigation is logical and clear.
c) Give more significance and publicity to the system.
The system has been promoted within the search engines and with many of the
sites featured by the system. The email base has been extended, but entries in
directories and general publicity appear limited. Some National nodes are far
more active than others in promoting the system.
d) A wide range of subjects is required
The subject range of the service has been improved in particular Natural
resources information and Socio economics material, see the breakdown of
Agriscout topics below:
Percentage use of topics in the Agriscout database


                                        Veterinary
                                           2%


                                                             Agriculture
                   Soc Econ                                    25%
                     25%




                                                                                    Fisheries
                                                                                       4%
                                                                                 Forestry
                                                                                    7%
                              Nat Res                                      Horticulture
                               28%                          Nutrition          3%
                                                              6%




e) To develop a news server
The news service is unique in providing an ARD news service, with general
services such as Northern Lights6 and Moreover7 having virtually no European
coverage in this area. It is recommended to issue the News service periodically as
an email to attract more awareness of the service and encourage registration to
the site.
f) Target the main user groups


6
    http://www.nlsearch.com
7
    http://www.moreover.com



                                                                                           16
EIARD Infosys phase II Assessment report                           C Addison        February 2001


Whilst the user study recommended addressing the different users, Research,
information providers and those involved in education with different interfaces,
the core group considered this was not worth pursuing. This has also applied to
the idea of different language interfaces.


4) Functional structure as a decentralised system.
The functional structure of the system relies on decentralised input to a number
of indexes.
These are clearly identified for the following services: News, Events, Discussion
fora, training courses, experts, Funding opportunities, Networks
In the case of Institutions, Projects and Metadata collections although InfoSys
provides a central index for entry, substantial data is available in other systems.
This has lead to the development of a multisearch tool to present these data sets
as one to the user.
The database of databases e.g. the metadata index, points to sites for additional
Projects, Publications, news services, Job Vacancies and whole Information
Systems.
The drawback with this method of decentralisation without central review of
content means that the quality of the content varies. This is particularly difficult
where the country participation covers such a variation in scale. Whereas Ireland
may have 20 individuals concerned with Agricultural Research for Development
Germany has 4000. Therefore individual websites for researchers will be entered
by one Node and not by the other.
In particular several participants in InfoSys have raised a concern over the scope
of resources entered in the InfoSys system, in particular within Agriscout (The
metadata resources). The definition of ARD has been interpreted differently by
different groups. The most important of these is the differentiation between
including research in similar agroecological zones in the North, and only including
research from overseas.
The best approach is a hybrid where the systems are decentralised as far as
possible whilst maintaining a central review or management function in the
partnership. Most systems use a combination of end user submissions and data
entry by nodes.


5 Future prospects for InfoSys
Looking forward at prospects for the InfoSys system needs to take account of the
possibilities for this combination of decentralised services with decentralised
input, linked and featured by an Agricultural Research for Development Window.
There is a clear value to coordinating the production and indexing of materials as
a European group, but for much of the contents need to be featured and
promoted in a global system. This may point to the increased role of InfoSys as a
network of information providers and managers and providing the ARD view on
European data to provide to other services for example the Electronic GFAR
system (EGFAR).


5.1 Prospects for EIARD-InfoSys
InfoSys should focus on a user search assistance service. Hopefully identifying
frequently asked questions and advising the ARD information management
community how best to answer them. In particular it needs to serve the needs as
expressed by EIARD working group8. Whilst it can never provide all the answers

8
    Questions posed to InfoSys by the 19th EIARD WG meeting, December 2000, Paris



                                                                                                17
EIARD Infosys phase II Assessment report                  C Addison    February 2001


to the types of questions posed by the EIARD working group directly it links to
sites that can. It could provide guidance through frequently asked questions and
special views, e.g. partner selection and funding, on how to answer specific types
of query.
This may involve developing new services but is more likely to involve pointing to
existing services and encouraging those providers to produce an ARD view on
their data. This has begun with links to services such as the Training database
run by Infoagrar.
It would be sensible to extend this approach to entry of certain data in the
international directory. For example in the case of crop specific information it is
worth considering featuring and promoting the Open directory, particularly as this
is now used by the major search engines9. Google integrates this in it's main
search, displaying the relevant section of the directory if it matches the user
request.
The data entry module could be developed and redesigned to provide a general
service for web promotion rather than an input solely for the benefit of the
InfoSys site. Links to the data entry pages of the Open Directory and Agricultural
directories together with a link to the Oneworld pages on how to promote your
site (for development organisations) might be an interesting site.
This does not need to involve major effort and could be spread amongst partners
who already have substantial experience in this area.


5.2 Options for the InfoSys strategy
The conclusions from the SWOT analysis10 undertaken prior to the Core group
meeting raised a number of options for the InfoSys strategy11 which were
developed in discussions during the planning workshop. These focussed on the
service strategy for InfoSys and the maintenance requirements.
InfoSys must serve the needs of the EIARD community as represented by the
Working Group. The emphasis must therefore be on providing two clear services,
namely Research Partner Finding Services and Funding Views. Both these
elements were developed in the Policy discussion group at the workshop.
Both these options illustrate the present advantages and limitations of the
service. Whilst research partners can be found through the various Organisation
databases this requires guidance for the user and the presentation of related
information through a new view of the data.
The other clear conclusion is the need for the network supporting EIARD InfoSys
not only to be based on National nodes but also having a closer relationship with
the associated networks and the National Fora for ARD. This broader network
might serve the communication needs of InfoSys as well as Information indexing
which has been the present emphasis.
Any decentralised service is very reliant on completing the development and
operationalisation of the Multisearch tool. This is particularly the case for Project
and Organisation information.


5.3 Does ARD in Europe require a specialised information system
The present EIARD InfoSys provides several functions, some of which are also
served by International systems. The focus needs to be on those areas where


9
    See http://www.google.com
10
     Annex 1 SWOT analysis
11
     Annex 2 presentation at Core Group



                                                                                    18
EIARD Infosys phase II Assessment report                  C Addison     February 2001


InfoSys has added value. Information on events, news Organisations and experts
will be particularly useful in a region where possibilities for collaboration, meeting
and funding are relevant. Policy information relevant for Donors will also have
European specific aspects so would also need to be covered. Resources which are
more scientific in scope and subject rather than location specific fit more readily
into a global system.
Even with the global systems emerging there is still a need for a network of
information providers to ensure the coordination of presenting and promoting
European ARD information. It should be clear that the InfoSys window is one part
of a broader European activity through various services of identifying and
indexing ARD material. It is therefore useful to have a platform for ARD
information providers.


5.4 The InfoSys niche
In determining the potential future role and niche of InfoSys interviews were
carried out with other networks. The discussions all show a potential continuing
role for coordinating information management at a European level.
Search engines
The generally accessible search engines for example Google, Yahoo, Altavista and
others are extremely powerful for specific terms. Hence in Agricultural Research
for Development they are very good for identifying research resources where the
latin name of the crop, animal or pest is given in the document. This avoids
ambiguity and only relevant texts are retrieved. If however you are searching for
a concept like plant processing in such a search engine, retrieval is far more
complex. In InfoSys you will be presented with a page of results that are directly
relevant to Plant processing in the ARD definition. Such full text systems are of
course also limited in finding partners in Europe working on ARD for example.
Research for Development
Services such as Euforic and OneWorld provide access to very different materials
on ARD. In both cases, either through the library in Euforic or the full text search
in OneWorld you are linked to individual documents. These documents only
provide a partial view of ARD. Euforic documents are restricted to policy relevant
observations and have few recent documents on ARD. OneWorld has a larger
coverage of NGO material rather than scientific research, due to it's membership.
Results will therefore often be drawn from NGO sites and again be related to
policy or campaigning.
Project and Organisation related information
Wisard provides the services behind other project systems such as IPMEurope and
provides an easily navigable system for identifying Projects and Organisations.
Like the ETFRN organisation database this provides more comprehensive
collections for specific subject areas such as IPM. The InfoSys system has recently
linked with ETFRN to provide searching possibilities within the Agriscout
organisations section in InfoSys. There are plans to extend similar services with
WISARD. At present there is some overlap of entries, but there are a number of
project databases with similar content. By reaching agreement with CORDIS,
ZADI and IAO for Project information to be cross searched WISARD data could
also be featured through the InfoSys site.




EGFAR and Regional Information Services




                                                                                    19
EIARD Infosys phase II Assessment report                          C Addison      February 2001


The Global Forum for Agricultural Research pointed to progress with the Regional
Agricultural Information services in Latin America and the Caribbean (Infotec12)
and the Asia/Pacific Agricultural Information System (APARIS13) both follow
similar approaches to InfoSys.
The management of an Information system such as InfoSys requires input and
support from both the users and those involved in technical delivery. The systems
work where
a) End users who are interested in the tools and experts in the process can act as
brokers between the system and it's intended audience.
b) This also implies that attention should be paid to improving the capacity of the
end user to apply the system and implies a change of attitude as well as skills.
c) The change of attitude means a mentality for sharing information more readily
and an appreciation of the global environment.
GFAR is particularly interested in constructing a knowledge pool and working with
pilot projects to adapt this idea.
The quality of an active system could increasingly be maintained by the user
through their use of items and feedback and criticism. This would lead to a
dynamic marketplace with built in mechanisms through which end user react to
poor quality. This would use for example the type of ranking of results employed
by Google where those that are selected most often appear first.



6 Technology

6.1 Extensible Markup Language (XML) and InfoSys
The Web is currently undergoing a transformation as more developers and site
managers change the way their data is stored and presented through the web.
The quiet introduction of XML behind the scenes has already provided us with new
features and services on the web. This transformation is the subject of special
project within the World Wide Web Consortium (W3C).
The semantic web is a project to examine how the contents of the web can be
split from the way of retrieval and presentation. It relies on the fact that all data
can be saved on the web in a markup language which describes the data
elements, together with a separate file, a stylesheet which will determine how it
is displayed.
These xml files, which describe the data within tags, still require some
standardisation to be useful. Therefore for a document you may have the author's
name embedded between the following tags <AUTHOR></AUTHOR>. It is easy to
draw a series of these files together from different sites on the web and
assembling them into one larger file. This is then displayed to the user through
the associated stylesheet. The stylesheet may be a simple list that just displays
the author's name, or it could display all the details in the file.
When running a site like InfoSys this would allow many more features than are
possible at present. In generating the news, items could be taken from other sites
and as long as they were in the same format could be easily incorporated. In the
same way by generating the news feed in a standard form it could be provided to
other news services.

12
 Infotec Regional Agricultural Information System of Science Technology and Innovation,
FORAGRO/Technical secretariat supported by CRIIA & GFAR
13
     First APARIS Activity Report December 2001



                                                                                             20
EIARD Infosys phase II Assessment report                 C Addison       February 2001


By adopting standard lists of tags for XML files for projects, organisations and
other information types InfoSys can exchange, combine and broadcast the
information elements in a common display. Whilst all this is potentially possible it
is important to focus on what the user needs and what is affordable or feasible at
present.
The current EIARD-InfoSys approach of developing an XML (wrapper) ie a
program that runs on the server to generate XML files, is an interesting approach.
With this and the gateway software a small number, say 5-6, of databases could
be searched dynamically and the results combined in one results page. This would
be suitable for the present project databases and organisation databases.
For larger number of databases the results would be very slow and so the data
should be collected before the user requests it. This would be possible with topics
or news pages where data from other sites could be included on the same static
page. This is elaborated in the next section.


6.2 Multisearch possibilities
The user cannot search the contents of all the information services featured in
InfoSys. To allow complete searching a multisearch facility is being developed but
it is important to understand the scale of searching.
The multisearch can be at a variety of levels each increasing in cost.
a) Links to systems user searches from search page one system after another.
Present situation.
b) Links to systems with search. User searches for topic is presented with relevant
databases and by clicking goes direct to search results on that topic. This is used
by ETFRN in their topics pages.
c) One Search page for many databases. The user has one search page and then
browses through database results by clicking on each relevant database name.
ELAND uses this approach in the Navigation bar.
Results from several databases are combined in one page. This may be achieved
in several ways.
a) By preparing a static page with results from databases glued together (XML
allows this to be done in a common format and to sort the results using a
stylesheet) If the results are in HTML they are simply displayed one after the
other.
b) By a dynamic search where the user issues a search and results are returned
from relevant databases and combined in either of the ways described above.
c) By a search on a database containing the entries from all databases in the
system. The advantage being speed for the user, the disadvantage being
management of the central resource and ownership issues. This central database
could be in two forms:
       i) A simple spidered collection (A search engine like google). This would
       copy whole records as full text and allow you to search in this form. Types
       of database could be separated by classifying the databases when they
       were spidered (collected).
       ii) A structured collection. This requires that the data is transferred in a
       common format so that individual fields in the database are matched, the
       huge advantage being that individual fields such as country can be
       searched.
In both cases retrieval is limited by the vocabulary used in the database.
Countries may be in a different language, dates may be in a different form. For
this reason standards are required if the search is to have a higher value.



                                                                                     21
EIARD Infosys phase II Assessment report                    C Addison    February 2001


In reality the two approaches exist. For example Google use a full text search
service but if it also finds an entry in it's structured database it presents a link to
this.
Google uses the Open Directory as it's manually indexed database. This directory
is open to the public to submit articles. The only quality control Google adds is
how many times the link is clicked on by the user.



7 Sustainability

The sustainability of InfoSys as a service is far from just a question of money. If
the inter-reliance of the various services can be increased the sustainability of the
InfoSys site will be ensured. For example the networks and associated partners
are paying for platforms individually. If they can start sharing platforms the
central costs of InfoSys will reduce dramatically. This depends on individual
organisations and sectoral networks as much as the National platforms.
This happens already with IPMEurope for example using WISARD as its platform
for projects. All the Agriculture related networks could take advantage of the
InfoAgrar Training courses database. The events calendar could be used by the
different networks if they are given views that allow them to display only their
events in a window on their site.


7.1 Devolution of InfoSys functions
Considering the present scope of InfoSys it would be difficult to devolve the
coordinating function to any of the existing systems. Whilst many offer partial
coverage of the services none would supply all the functions. The key elements of
the service are:
a) The website - Navigation to the constituent databases
b) The network - Organisation of Core group meetings dialogue with project
partners, development of joint projects.
c) The databases -     Agriscout (metadata)
                       Events
                       News
                       Funding
                       Experts


Some InfoSys functions would lend themselves to devolution more than others.
WISARD might form the closest match for some of these functions in relation to
Project and Organisation data held in Agriscout, but this would still need
maintaining and as mentioned elsewhere project data is already likely to rely on
devolved databases through the multisearch software.
Databases in the other areas could be devolved if there was a specific interest but
this will complicate the data input. At present through one login the various
databases can be updated, if they were all to be devolved there would need to be
agreement on common indexing and ideally login access, this would involve
significant costs to establish. It may be possible to continue a central login
system which allows access to the various update systems.
There are no obvious candidates for running the other services as the associated
services have different mandates. CTA is not considering running European



                                                                                      22
EIARD Infosys phase II Assessment report                   C Addison    February 2001


services, although they would be interested in looking at email links to ACP
partners and working with dissemination of information to ACP partners.
Devolution of the services must be balanced against preserving an identity for
EIARD InfoSys and not diluting it to the extent that it is seen as several systems
rather than one.


7.2 Membership models and fees
The following models were presented to partners in the project, but there was no
support for directly contributing to the European core activities.
Euforic - a cooperative based on Information providers building a common
platform to present policy relevant information. Europe's forum for International
Cooperation.
This is based on a small secretariat of 2 staff with extra project staff (up to 2)
running two main databases, a library and an organisations database.
These are supported through a series of projects and services with membership
being less than 20% of the income. The service is heavily reliant on projects and
services and it is consequently difficult to have an independent strategy and to
ensure long term stability.
ETFRN- European Tropical Forest Research Network, a network of European
researchers with an associated website providing access to existing resources.
Small secretariat of 2 plus part time web manager.
Although there are a large number of registered members (over 2000), they do
not pay membership fees. It seems unclear how many would stay if charges were
introduced. The advantage of the registration is clear interest and feedback from
users. Support for activities is through core project funding and holding joint
events and through joint publishing with partner organisations.
In discussions with InfoSys project partners it is clear that National membership
would not be able to fund central activities. Views vary because of the vast
difference in numbers of those working in ARD and the relative importance at a
National level. More likely was membership at an Organisational level. This would
be of interest if linked with services or the possibilities to promote certain
services. This would have to be handled carefully as many of the project partners
were nervous of changing the balance of participation and including a much
higher number of private sector organisations.
The main drawback in collecting membership fees are the costs incurred
compared to the amounts organisations would be prepared to spend. It does not
seem feasible to raise the core costs just for maintenance (about 100,000 Euros
per year) as Organisations only seemed prepared to pay a maximum of 2000
Euros per year for membership, with smaller groups expecting banding of
charges. There was also some concern as to how membership could be inclusive,
particularly as smaller NGOs might be interested in joining but not be able to
afford membership.
In summary it seems impossible to establish a paying membership network in this
area. Income could be gained as now through a series of projects ideally jointly
with project partners.


7.3 Establishing a donor group
Whilst this may become a future possibility the present InfoSys would need to
more directly serve the EIARD requirements. However as a first step the Working
Group should consider sharing information on current donor support to ARD
related information services, this would illustrate the current level of support to




                                                                                     23
EIARD Infosys phase II Assessment report                          C Addison   February 2001


this activity by donors and the necessity for some expenditure on coordinating
this effort.


7.4 Network of excellence
There is clear support from many of the InfoSys partners to participate in and
support the ARD network of excellence14 concept proposed for the new framework
programme. The strength is that the network provides a basis for support and a
consortium of information providers is already in place. The weakness is that the
new multisearch technology is not yet proven, but would be in place before the
proposal was considered.
The approach linking a new gateway function for European ARD resources with an
e-University has shown significant interest and support at International and
National level. The proposal is seen as coming from a qualified consortium from
the training and information perspective.
The multiplier process of training of trainers and the encouragement of partners
to develop their own information systems along common standards for integration
and retrieval.
This would appear an appropriate use of the information system in support of a
capacity building initiative in developing countries.




14
     EFARD proposal for the ERA-ARD Annex3 Fiche of the proposals 18



                                                                                          24
EIARD Infosys phase II Assessment report                C Addison   February 2001




Glossary
AgriScout - Metadatabase index of Infosys site
AIDA - Project database of the Development Gateway drawing on donor
databases
APARIS - Asia Pacific Regional Agricultural Research Information System
ARD - Agricultural Research for Development
Asia Urbs - European Programme for Urbanisation projects in Asia
CORDIS - Project database of European Commission funded research
EGFAR - Electronic Global Forum for Agricultural Research
EIARD - European Initiative for Agricultural Research
ETFRN - European Tropical Forest Research Network
EUFORIC - Europe's forum for International Cooperation http://www.euforic.org
GFAR - Global Forum for Agricultural Research
Google - Search Engine - http://www.google.com
IAO - Istituto Agronomico per l'Oltremare,Florence, Italy
InfoAgrar - Web gateway for Agricultural Research for Development information
in Switzerland http://www.infoagrar.ch
IPM Europe - European Integrated Pest Management Working Group
Northern Lights - Search engine and News Service http://www.nlsearch.com
OneWorld - Web Portal for Human rights and Sustainable Development
http://www.oneworld.net
SWOT - Strengths, Weaknesses, Opportunities, Threats
W3C - World Wide Web Consortium
WISARD - Web-based Information System for ARD
XML - Extensible Markup Language
ZADI - German Centre for Documentation and Information in Agriculture, Bonn




                                                                                25

				
DOCUMENT INFO