Błażej Feret, Michael Kay

eIFL - Electronic Information for Libraries a global initiative of the Soros Foundations Network

with co-operation from Anna Maria Balogh, Peter Szanto, Sam Brooks5

---

The text of the talk presented during the 67th IFLA Council and General Conference, August 16-25 2001. The original text avaliable online at:
http://www.ifla.org/IV/ifla67/papers/117-141e.pdf

On September 1, 1999 the Open Society Institute - Budapest (OSI) and EBSCO Publishing agreed to collaborate in a project to create the largest information consortium in the world. The joint multi-million dollar initiative, called Electronic Information for Libraries Direct - eIFL Direct, makes information available electronically to libraries (academic, research, medical, public, national, and parliamentary), ministries, and non-governmental organisations (NGOs). Users include scholars, policy-makers, government officials, students and the general public in countries in transition. Over 2 000 institutions now belong to the consortium and the number continues to grow.

To date well over 3 million searches have been performed and this number is increasing markedly as more institutions join and users become familiar with the service.

The venture targets traditionally "information-starved" countries - these are the countries primarily supported by OSI, whose mandate as part of the Soros foundations network is to foster the development of open societies. Access to the widest possible range of information is obviously a major factor in fostering this development. Along with Guatemala and Haiti, 27 countries are from Central and Eastern Europe (mostly former Soviet satellites), and ten are from Southern Africa. The project will be expanded in the coming months to 18 more Western African countries and Indonesia enlarging the consortium to 58 countries.

Evaluation of needs and offers

After a preliminary evaluation of the information needs of libraries in the region, OSI issued an open invitation to tender to all companies dealing with electronic journals particularly in the social sciences and humanities. A total of seven companies submitted proposals. Experts in the field from OSI reviewed these, and three companies were short-listed for further consideration. Two professional teams - one based in Russia and the other in the United Kingdom - then independently evaluated the products. The teams were given over 15 main criteria and 28 sub-criteria to use in their evaluation. The unanimous decision was that EBSCO provided the most comprehensive service available on the market at that time. The OSI Information Sub-Board and finally the OSI Board endorsed this recommendation.

The full list of criteria used for the comparative evaluation of the provider of full text "soft science" (social sciences, humanities, management and economics) are listed in appendix A. Given OSI's mission and the range of countries it serves, the evaluation criteria also included reference to the number of full-text journals with specific focus on Central and Eastern Europe and the former Soviet Union as well as Africa. The need to deal with countries with rather poor connectivity was reflected in the specifications regarding the supply of CDROMs. The selection criteria dealing with local content ("Adding Local Content in English" and "Adding Local Content in Local Languages") already pointed to one of the directions the eIFL project would take in the medium-term future. That is to become an umbrella for East-East and South-South delivery of content of local relevance and in languages other than English.

Product enhancements since the launch of eifl direct

At the time of the tender EBSCO was providing just under 3,000 full text titles through the EIFL project. Since the project began, over 2,000 new journals have been added to these databases, bringing the total number of available full text journals offered in the eIFL Direct suite of databases to over 5,000. The focus has also shifted from almost entirely soft sciences journals to a mix of soft sciences and STM journals. There are now over 1,000 hard sciences (STM) journals available in full text in the databases available through the project. The Academic Search and Business Source databases have been upgraded to Academic Search Premier and Business Source Elite.

At the time of writing eIFL/EBSCO offers access to six larger databases at the same fixed 3 year cost:

• Academic Search Premier (upgraded from Academic Search Elite)
• Business Source Premier (upgraded from Business Source Elite)
• The Masterfile Premier
• Medline with Biomedical Basic collection
• Health Source Plus
• Newspaper Source

EBSCO is currently working on a Russian interface, which will be particularly useful for many of the users not just in Russia but also in the CIS. This should be completed within the next six to twelve months. Further, EBSCO is in the process of licensing journals in Russian. These journals are expected to be available to the eIFL consortium at some point in 2002. Meanwhile titles published in English from Central and Eastern Europe are continuously being added to the database.

Expanding to "hard science" content

The formation of the Content Task Force

It was clear from the beginning that there would be some start-up difficulties in a project of this dimension and geographical complexity. In the second half of 2000 the continuation of the eIFL project was threatened by the reality that not all countries had been able to identify the necessary funding. This nearly resulted in a classic Catch 22 situation: libraries were unwilling to find funding without a clear commitment to the continuation of the project - but this commitment would only be given if there was some guarantee about forthcoming funding.

Within OSI there was no doubt however about the two main 'saving' principles:

• there had to be full involvement of the stake-holders, the libraries, if the project was to survive and prosper,
• this involvement had to be based on the consortium approach as this form of organisation is best suited to deal with the issues faced by libraries when licensing electronic resources,

This scenario led to the decision to constitute a Task Force on Content, to guide future expansion of subjects and titles offered under the eIFL umbrella. It now meets regularly to evaluate services and content offered by EBSCO and to consider adding new content.

The first priority of the Task Force was to deal with the demand by the libraries in the eIFL project for STM (with the stress on ST) content. This was in response to an extensive survey of the additional information needs of the consortium. (see Appendix B) The aim is to provide this traditionally highly expensive information to countries of the consortium at equitable prices and based on any individual country's state of development and ability to pay.

The first problem to be solved by the Task Force regarding the science and technology content was to decide how the choice of full text resources should be made. There are several methods practised in libraries:

• Extending the current print subscription contracts to include full text online. Usually there is no research, no surveys and no evaluation before making this choice.
• The library doing its own research. The library defines the selection criteria and conducts tests and comparisons of the available products.
• Basing any decision on the opinion of advisory or consulting companies or bodies. A good example is "The Charleston Advisor" http://www.charlestonco.com/ , which devised a rating system scoring each product by the following elements: content, searchability, price and contract options/features. The company regularly publishes reviews and comparisons of web products. Another example of an advisory body, which carries out and publishes studies on the evaluation and analysis of full text databases is California State University Libraries Electronic Access to Information Resources Committee (CSU Libraries EAR) http://seir.calstate.edu/ear.rev.fm.html/
• Asking users (via surveys) before the choice is made, what products they want (based on the opinions and evaluation of the users). Here the usual practice is to provide users with a list of services or web links and asking for the evaluation of each of them.

It is common practice in libraries to take a decision regarding which services to contract in one (or a combination) by one of the methods described above. The selection made is later confirmed (or rejected) via surveys distributed amongst library users in the form of database questionnaires or evaluation forms. Examples of such forms may be found at http://library.sjsu.edu/test/testeval.htm or http://www.library.unsw.edu.au/exit.html.

The Content Task Force considered two options for how the selection procedure could be performed:

• analyse the products on the market, evaluate them according to the predefined criteria and select those which the best "fit" the criteria,
• ask librarians (and/or) library users what products they would like access (which products would be essential, or at least useful, for their users).

A third option was proposed and this was in fact the preferred model but considered too difficult to implement. This was to analyse the use of separate journal titles across libraries and then to negotiate access to the full text with the individual publishers involved. This would be a logistic nightmare given the size and diversity of the eIFL consortium.

Finally the Task Force decided to prepare an evaluation survey, in which a large number of services, along with a brief description and URL, were listed (see Appendix B). Respondents were given three choices: 'essential', 'useful', 'not needed'. Libraries participating in the survey could also add other services, which are currently used or would be useful in the future. It also listed over 20 bibliographic databases that could be made available through eIFL. A similar approach has been used for example by Hungarian colleagues during the preparatory phase of a national science and technology consortium in 2000.

The survey was sent to the national co-ordinators of eIFL in March, 2001 with the request to distribute it further to existing participating, as well as non participating, libraries. At the time of writing, 199 libraries and research institutions from 19 countries had sent in their response.

The survey results cannot be published in this article as it was considered inappropriate for S and T publishers to see the results given the forthcoming tender.

• Many institutions desired access to a large range of full-text services.
• Whilst bibliographic databases were seen as important there was not nearly as much enthusiasm to have access to these as full text journals.

Based on these results, it was decided that:

• The first round tender for S and T should focus on full text alone and that a second round tender would then include bibliographic databases.
• Given the number of publishers and services identified in the survey, the tender would be completely open to the S and T publishing industry and that the Content Task Force would not pre-select a number of companies.

The tender process will commence in June 2001 and will be completed by January 2002. Details will be available on the eIFL web site http://www.eifl.net.

Towards the multi-country consortium

The OSI Information Strategy 2001 describes the principles of the eIFL project as follows:

"The project is based on leveraging two features of digital information. First, the Internet's network effect enables demand aggregation on a very large scale: a large consortium of individually poor consumers acquires significant negotiating power. Second, the marginal cost of digital data is zero, so that additional units of highvalue digital commodities produced for affluent markets can be resold to less affluent markets at negligible cost to the producer, as long as the transaction cost to the producer is kept low".

It describes its important future role in enabling East-East, South-South and South-East content flow, and in becoming a platform for developing local digital content.

Finally, "the consortium could become a bulk buyer of low-cost hardware and software applications for its member institutions, and an infrastructure for delivering training. At a later stage, it could begin to function as a network for dissemination of policy knowledge and participation in national and global information policy formation".

This is an extremely important strategic vision for the countries in the OSI portfolio. For the first time in many countries, it will facilitate a powerful platform for joint stakeholder action, giving a voice to the library community - a community, which plays a fundamental role in the development of an open society.

The future mission for eIFL will be to lead, support, motivate, and advocate for multinational library cooperation among member libraries for nations in transition around the world. It will develop and deliver innovative library programs and services that advance collaborative activities, resource sharing, and other programs that are responsive to the needs of the library constituencies.

Appendix A

Criteria for comparative evaluation used when choosing social sciences full text provider:

Titles
Total number of titles in full text.
Total number of titles providing abstracts only.
Total number of titles in full-text excluding leisure and STM journals.
Total number of abstracts excluding leisure and STM journals.
Total number of publishers contributing.

Breakdown by Subject Areas
Total number of full-text journals by the following subject areas: Economics, Education, General social sciences, History, Law, Philosophy, Politics, Psychology, Sociology.

East vs. West titles
Number of full-text journals with specific focus on Central and Eastern Europe and the former Soviet Union.

Overlap of titles
An analysis of the percentage overlap between full-text titles from the two services being evaluated.

User interface
An evaluation of the user interface including user friendliness, level of searching i.e. basic to advanced, help screens, any other features.

Special Modules
Any special modules e.g. usage statistics etc.

Terms and Conditions
Sample contracts were reviewed by attorneys.

Access details
Are there any limits to the number of simultaneous users for both the online and CD ROM versions?
Is access only available in the library or it can be from the university sites on the same campus or elsewhere?
Is access possible from home?
Any other special technical requirements to access the service?

Response time
Assessment of response time testing the system at regular intervals during the day.

CD vs. Online
Will CD be provided for every site in the country?
How often are CDs updated?
Can CDs be networked or they are stand-alone?
Are CDs offered for all products?
Are all countries eligible to have CDs?

Price Analysis
Excluding leisure and STM titles, what is the cost per current title?
Excluding leisure and STM titles, what is the cost per title including all issues assuming average back-titles go back three years?

Training and Support
What is the training strategy and is it free?
What helpdesk support is provided?

Professionalism and Pre-sale Service
What was the level of professionalism and attention to detail that the companies showed when dealing with OSI staff and the independent evaluating teams?

Adding Local Content in English
What level of commitment do the individual companies demonstrate to the idea of adding English language titles with specific relevance to the region?

Adding Local Content in Local Languages
Are the companies prepared to add local language titles in the medium term future?

Appendix B

Name of institution
 
.........................................................................
 
1. Which of the following STM bibliographic databases would your library wish to have access to? If you already subscribe to any of these databases, please indicate which ones.
 
Please add any other that tou currently subscribe to, or wish to subscribe.
 

 

 

 

2. Which of the following full-text science and technology electronic journal services would your library wish to have access to? If you already subscribe to any of these services , please indicate which ones and what your current annual subscritpion is in USD.
 

 

 

 

 

3. Which ministry(ies) in your country does your institution receive funding from?
...............................................................

4. Do you receive any funding support from other sources such as private sector sponsoring?
...............................................................

5. Would your institution accept private sponsoring to subsidethe cost of access to these services if funds were avaliable? ...............................................................