Minutes Elisad Extraordinary Gateway Group Meeting

Stockholm, Wednesday 24^th October 15-18.30

Present:
Anne Singer (ELISAD)
Thomas Rouault (TOXIBASE)
Marie-Lise Priouret (TOXIBASE)
Patricia Brigoni (Gruppo Abele)
Stephan Schulte-Nahring (DrugScope)
Gunnel Larsson (CAN)
Marianne van der Heyden (Bureau Andromeda)
Toine Ketelaars (TRIMBOS)
Susanna Prepeliczay (ARCHIDO)
Apologies:
Hana Sovinova (SZU)

Minutes:
Stephan Schulte-Nahring (DrugScope)

Agenda:

1. Project Working Plan : Involvement, responsibilities, coordination of tasks of participants

- distribution of project guidelines for scope policy, selection criteria and collection management (ARCHIDO/BISDRO)

- short report from partners on estimated amounts of websites for the gateway data collection (FR, CZ, UK, SE, NL, IT, DE & other countries) and use of abroad contacts planned

- planning of an editorial board (TOXIBASE)

2. Gateway data collection tool: Conceptual reflections on new evaluation form version (prepared by ARCHIDO basing on ELISAD, DrugScope, BISDRO input)

3. Gateway website: Conceptual reflection of sitemap draft in comparison to other gateways (addictionsearch, SOSIG, OMNI) (prepared by ARCHIDO)

Topics discussed

The discussions during this meeting didnt strictly follow the agenda points as listed above. Agenda point 3 was discussed on the second day of the working group meeting (25^th October 2001). Susanna first reminded working group members of the subject gateway project background and its objectives and informed the project team that the European Commission (EC) in June 2001 gave a principal approval to fund the project. Most of the time was spent on discussing issues to do with funding, subcontracting, Susannas role as project administrator, quality control, what is to be evaluated, countries covered by the gateway and aims of the gateway project. Susanna allocated the estimated numbers of websites for different countries that are to be evaluated by working group partners at the end of the meeting.

Starting date and monitoring of work done

Decision: It was decided that the official starting date of the subject gateway project is the 31st December 2001. Thomas said that each partner involved in the project needs to monitor and document the time spent and work done in order to be able to provide proof of his/her contribution in kind. The two subject gateway project meetings in March 2001 and October 2001 do count as a contribution in kind but should not be given a date because they took place before the official starting date; all work done before the official starting date needs to be declared as having taken place after the starting date.

European Commission funding for the subject gateway project

Susanna started off with a brief reminder of the subject gateway project background and its objectives and that the EC in June 2001 gave a principal approval to fund the project. The EC will pay 70% of the costs of the project.

The 70% amount to EURO 158.634. The total project costs are EURO 226.619. The remaining 30% represents the contribution in kind which amounts to EURO 67.985.

The European Commission payments will be given in three stages: an initial payment, an interim payment and a final payment.

Two payments will be paid out to gateway project participants: one interim payment and a second payment that will be possible only after the final accounting work with the EC has been done because the final payment will be paid AFTER we have proved that all costs have been invested correctly. This means that we might have to give in advance evidence of costs that the EC payment can only cover at a later stage after the work will have been done. Therefore it will be necessary to have relevant letters, documents such as invoices / bills etc. for the EC administration. Participants will be asked to submit papers for administrative purposes to meet any EC demands (also before having received the final money). The point was made that we have to distinguish clearly between the project content work and the administrative work. We will need to clarify who does what for the project. Susanna will do a listing of the tasks which will have to be agreed by the partners involved.

This is especially true for the data collection part of our work. The subcontracts will show work amounts according to the rough estimates (websites listing) and in the process must include the possibility for adaptation to the actual number of evaluated websites.

Susannas (ARCHIDO) role as project administrator

Marianne suggested that ARCHIDO will have to act as a co-ordinator of the project content work and organise the administrative processes. Susanna specified a wide range of tasks that she, Marianne and Anne need to do.

These are:

Allocation of data evaluation work to project partners and co-ordination of subcontracts

allocation and distribution of the tasks of the 3 persons employed in the project for 2-4 months: Marianne, Anne and Susanna

Their tasks will be:

website planning, layout/design and creation, publishing of HTML documents, planning and design of search forms and input forms, promotion, production of promotional materials, online questionnaires for data collection and project evaluation, user access statitics, report writing (2 reports: interim and final report)

organisation of the whole methodological background work

project development and process co-ordination: contacting/mailing, time framework, generally allocating any other tasks

output drafts (drafts of search result reports: one simple form and one detailed form)

? ordering the structure/content of the website (Susanna, Anne and Marianne develop the sitemap draft)

? browsing list drafts (NOTE: each browsing list will have a related thematic heading and sub-headings within the topic area. These lists display the title and links as search results. A logical structure and an internal linking of related thematic issues / content overlapping need to be developed)

meeting planning (organising meetings and drawing up agendas)

technical realisation: database conception and implementation, server organisation, software functionality, search facility in cooperation with the programmer; central database content management

organise and collect the work done

budget management and communication with the EC

Subcontracting

Thomas said that other Elisad members who arent involved yet can still join the gateway project through subcontracting, e.g. description of websites in other countries. Susanna responded saying that it should be made clear to all participants that a lot of the work is done on a voluntary basis but that some tasks can and probably have to be paid for. An example is the production of translations such as translations from Greek into English.

All members of the working group are to be allocated at this meeting a certain number of countries and a certain number of websites for each country that need to be identified and evaluated. In terms of procedure, if the working group partner cannot identify or evaluate websites themselves because they lack the language skills, they will have to first ask an Elisad member who is competent to do so, to identify websites and produce website descriptions on a voluntary basis. However, if that partner is not willing or able to do this work for free, then that partner or someone else who is qualified could do the work which is then being paid for by money that come out of the project budget. As this budget covers only a part of our work (70%), we ought to ask ELISAD members for voluntary contributions as much as possible. On a related matter, we can involve organisations that are not ELISAD members but that want to be listed on the gateway to fill in the online form. Anne has started a list of those who are interested. Susanna will send the data collection tool to ELISAD members via the mailing list so that every member gets the feeling that they are involved. This will happen after the development of the form has been completed. It was noted that the more project participants we have, the better for us because if one partner is overwhelmed by the number of websites (differs from country to country) it will be difficult for him/her to evaluate them all. Thomas suggested that we need to pay members to fill in forms if they do lots of websites. Susanna suggested that she could modify the budget to allow being able to pay partners.

Decision: Susanna will develop a structured table of data collection tasks and

explore possibilities for budget modification to allow payments for partners.

Quality control:

Marianne and Marie-Lise pointed out that we have to evaluate websites and validate other peoples input. Marianne suggested that we could use the funding to evaluate the gateway. The project plan and the projects budget make provision for an evaluation of the gateway by its potential audience. We ought to find persons with appropriate language skills who can translate texts into proper English. Susanna said that the EC wants us to only involve EU member states as paid partners in the project. She suggested that in order to simplify the project and for the sake of future maintenance and updating of the gateway, we ought to focus on Elisad members who will be able to evaluate websites and update them in a few years time.

In terms of the projects feasibility and with reference to the submitted project plan, Susanna also suggested that we could initially exclude those countries whose languages we dont understand or where we find it difficult to find someone who could translate texts into proper English. Susanna also suggested that the gateway coverage during the first 18 "official" months will be restricted to 18 European countries and pan-European / international sites (see table below: Included countries in the subject gateway during the first 18 months of the project)

These countries include some eastern ones where we have expert contacts (Elisad members): Hungary, Czech Republic. (who will also provide some data on Croatia or Slovenia) but apart from these, we will not broaden the coverage to include e.g. Romania or Poland etc. and/or overseas countries (US, Canada etc.) during the first 18 months of the project.

We will only be able to include other countries later on if it is feasible to do so. This would be where we have

a) experts with appropriate language skills and/or

b) resources in the framework of professional trainings and/or

c) extra funding for an extension to international coverage

Therefore, countries like Yugoslavia will not be included where we

a) dont have any contacts and

b) dont have working group participant with appropriate language skills

because this surpasses our capabilities. Within the next 18 months, we will only have to limit ourselves to European countries.

Marie-Lise asked about quality control and pointed out

a) the practicality of having secure contacts and

b) that we ought to make sure that the number of websites that have to be evaluated by our partners is manageable.

In terms of ensuring quality control, it was said that we need to make sure that the person who evalutes websites is an expert.

What is to be evaluated: website contents or the resources provided?

Stephan wanted to know to what extent an evaluator needs to be an expert and whether website contents is to be evaluated. This prompted a discussion about whether either the quality/validity of the contents of a website or the resources provided on a website ought to be evaluated.

Decisions: As a result of this discussion, it was decided that:

? it is not the task of the evaluator to check the validity of contents of a website beause there are very limited possibilities to do so. Even when checking the real" content compared to the claimed" content that is usually talked about in the about area " or mission area" of a website, there will always remain a gap between objectives and achievements.

It is the task of the evaluator to evaluate the information resources provided on

Websites according to the statements given on the website and according to our Scope Policy. Susanna reminded us that the subject gateway will have a

documentary focus and the target group is librarians and other professionals in the drugs field.

? we need a disclaimer which states that the listing of websites doesnt mean

that Elisad endorses its activities.

In order to tackle the points given above, Susanna proposed the following steps:

Step 1:

Test forms will be filled in initially by Elisad members (only a small number of websites) to have some test datasets that would enable us to test data entry, display forms/output forms and verify the areas covered by the questionnaire. Then relevant changes in the draft database structure can still be done before the distribution of the form to all Elisad members takes place. This doesnt necessarily mean the incorporation of such test datasets in the gateway.

Step 2:

Test forms will be filled in for websites across Europe.

Susanna will simplify the data collection tool (online form).

EMCDDA mapping of websites

The EMCDDA is currently doing a mapping of websites that cover the misuse of drugs; each REITOX Focal Point needs to produce a website listing. We could find out whether we could use the EMCDDAs mapping exercise as a resource for our own project.

Countries covered by subject gateway

Thomas suggested that we ought to include pan-European and international bodies. When a body has many branches, we need to make a decision as to whether we include branches or solely the head quarters (decision outstanding). Susanna suggested that Eastern European projects should be included as international bodies.

Decision: It was decided that we will include EU and international websites on the basis of the current group and find a reasonable amount of websites that can be handled by each partner on the gateway board.

Included countries in the subject gateway during the first 18 months of the project:

Susanna allocated countries and the number of websites for each country to the working group members (FR, CZ, UK, SE, NL, IT, DE & other countries) for data collection. This allocation process establishes which countries are to be included initially in the subject gateway over the next 18 months. All EU Member State countries are covered and some non-EU countries are included.

Decision: The countries and website numbers below were allocated to the following working group members:

Action points:

? Susanne will do a listing of the tasks which will have to be agreed by the partners involved

? Susanna will simplify the data collection tool (online form)

? Susanna will distribute the data collection tool to partners so that every partner gets the feeling that they are involved. This will happen after the development of the form has been completed

? EMCDDA website mapping task: we will find out whether we can use the EMCDDAs mapping exercise as a resource for our own project

? Susanna will change the to allow payments for partners

Minutes Elisad Extraordinary Gateway Group Meeting

Stockholm, Thursday 25^th October 9.00-12.00

Present:
Anne Singer (ELISAD)
Thomas Rouault (TOXIBASE)
Marie-Lise Priouret (TOXIBASE)
Patricia Brigoni (Gruppo Abele)
Stephan Schulte-Nahring (DrugScope)
Gunnel Larsson (CAN)
Marianne van der Heyden (Bureau Andromeda)
Toine Ketelaars (TRIMBOS)
Susanna Prepeliczay (ARCHIDO)
Guest: Dagobert Soergel (University of Maryland, SALIS)
Apologies:
Hana Sovinova (SZU)

Minutes:
Stephan Schulte-Nahring (DrugScope)

Agenda:

1. Gateway database planning :
- gateway dataset example (ARCHIDO)

- reflections on comparability of fields to international standards (EDDRA, DCMI, DESIRE) resulting from the post-Paris working phase (Toxibase, Bureau Andromeda, ARCHIDO summary) (not discussed)

- reflections on formats, utilities and tools (ROADS, server database properties) (not discussed)

2. Detection of website offers: Introduction to Internet Detective software (Bureau Andromeda)

The discussions during this meeting follow mainly the agenda points listed above although some issues werent discussed due to time constraints. Agenda point 3: Conceptual reflection of sitemap draft in comparison to other gateways (Addiction Search, SOSIG, OMNI) (prepared by ARCHIDO) on yesterdays agenda was also discussed today except for the OMNI example. The reflections on formats, utilities and tools (ROADS, server database properties) were not discussed.

Re: Agenda point 1. Gateway database planning : gateway dataset example (ARCHIDO)

Initially, the following points were made in relation to the data collection process and the form contents:

? audience of website:

The data collection form requires information on the audience of a particular website to be evaluated. Information about who actually visits a particular website can be taken from the guest book. Filling in this part of the form also depends on the personal judgement of the evaluator who is familiar with the contents of the website.

? thinking about how people can search the gateway

Marianne criticized that the data entry form was made before it was decided what the search screen of the database is supposed to look like. The approach to developing the gateway here first starts with input tasks. Marianne suggested that we should first think about how people can search the subject gateway before we design the data entry form.

? free text field for substances versus displayed controlled vocabulary

Susanna said that there are two options for searching:

1) free text searches whereby a user enters a word into a search field

2) field-orientated search where a limited controlled vocabularly offers user

search terms but doesnt allow users to enter free text

Susanna asked whether we ought to have a free text field for substances in which a user can enter the name of a particular substance and then searches the subject gateway on it. She suggested that we could offer field headers as labels for a particular free text field into which a user can enter keywords. This point was suggested as an alternative to having pull-down lists with a controlled vocabulary.

? making the form easier to fill in

Marie-Lise said that we ought to make the data collection form easier to fill in. She suggested that we ought to get rid of those fields which repeat sets of keywords. As an example, the following fields currently repeat the same set of names for psychoactive substances:

psychoactive substances information

substance specific focus of treatment

substance specific focus of service offers

substance specific focus of political activities

substance specific focus of prevention activities

substance (use) specific focus of research

Marie-Lise asked whether the amount of information we are currently asking for on the form isnt simply too much. She suggested that we should cut the form down in size and asked to be reasonable about the amount of information we want to collect on websites because we cannot anticipate every kind of information need that users of the gateway might have. She also suggested that we can be very descriptive in the summary description of the evaluation part of the form which requires a short general description of the evaluated website. Marie-Lise commented that full organisational contact details (part 1 of the current form) arent necessary. We should only give website address details. Gateway users who want to find out about contact details for a particular publisher of a website could follow the link to access such details themselves. She suggested that we dont have to describe an entire website, only essential bits.

Decision: we will not list the postal address of organisations that are authors of websites, just the website address.

Discussion of the technical implementation of doing searches using the gateway:

The comments about making the form easier to fill in prompted a discussion of the technical implementation of the search process. Marianne said that the data collection form and the data display forms are two completely different subjects which have confused the discussion to this point and suggested that we need to have a table to compare the input form with the output form(s).

Susanna said in reply to Marie-Lises point about the repetetiveness of the data collection forms contents that it isnt necessary to display all information that is represented on that form. It was only necessary for internal technical processes to have the large numbers of options and fields we currently have on the form but these will not all be necessary for data display.

Susanna wanted to save us from writing too much text. She therefore did not put many free text fields on the form because these require information to be filled in rather than simply being chosen (tick boxes). Stephan reported that DrugScope colleagues used the form to evaluate 3 websites and have overall found it clear and easy to use. The average time spent on evaluating one website was 30-40 minutes. His colleagues commented that some websites may be very targeted so that large chunks of the form were not relevant. So, despite the fact that the data collection form looks quite detailed, it didnt take much time to fill it in and many parts werent used because they did not apply. This is in contrast to websites with a very broad coverage of drugs, which require many of the tick options to be selected. Stephans colleagues commented that the most difficult and time-consuming part was the description which needs to be written in the summary field of the evaluation part. We then briefly entered a discussion on form details but it was said that we need to test the form first.

Decision: It was decided that the User Manual ought to define controlled vocabulary.

Decision: It was also decided that the summary in the evaluation part should describe the overall presentation of the evaluated website but not repeat whats already been said elsewhere on the form.

Re: agenda point 3. Gateway website: Conceptual reflection of sitemap draft in comparison to other gateways (SOSIG, Addiction Search)

Susanna proposed that we ought to learn from the design and presentation of other gateways that cover the misuse of drugs. During this session, the following examples of gateways were discussed: Addiction Search, SOSIG, and OMNI.

Addiction Search can be accessed at: http://www.addictionsearch.com

ADIN (Australian Drug Information Network), this gateway is co-ordinated by the Australian Drug Foundation (ADF) and can be accessed at: http://www.adin.com.au

SOSIG (Social Science Information Gateway) can be accessed at: http://www.sosig.ac.uk/

Discussion of the SOSIG example:

Susanna initially briefly presented some background information about this project. The Social Science Information Gateway (SOSIG) is a freely available Internet service which aims to provide a trusted source of selected, high quality Internet information for students, academics, researchers and practitioners in the areas of social sciences, business and law. It is part of the UK Resource Discovery Network. The SOSIG gateway only consists of the 8 fields. These are:

Name/Title

Description

Keywords

Subject Section(s)

Resource Type

Admin Email

Language

URL

The small number of fields (compared to the 89 fields on the Elisad data collection form) prompted a discussion about how we could make our data collection form easier to fill in and what information doesnt need to be collected. Susanna searched the SOSIG gateway using the keyword drug". It was said at this stage that not all fields on the database necessarely need to be displayed on the screen. The discussion of the Drugscope experience above suggests that not all fields will necessarely need to be displayed following a search. Susanna also noted that:

- SOSIG gives links to subsections of websites

- SOSIG has keywords that were assigned by the person who filled in the data collection form. SOSIG does not have a controlled pool of descriptors.

 

Discussion of what could represent added value

The small number of database fields of SOSIG records prompted a discussion of the benefits of having a short entry forms in contrast to having a detailed form.

The fact that the SOSIG form isnt as detailed as the ELISAD form suggests that we could make our form easier for data entry which would mean it is less time consuming to fill in. The danger was noted that this could effectively mean a loss of valuable information.

We discussed the example of a record on DrugScope which came up as one of the results of Susannas search. Stephan pointed out that this brief description doesnt adequately describe the Drugscope website and its resources. What was missing was an evaluation of what the databases cover in terms of contents, size and comprehensiveness. The description also lacked information about the types of resources available. This lead further to a discussion of what it is we wanted to get out of the subject gateway in the first place. Stephan reminded the team that the gateway was originally thought of as a resource for librarians and suggested that we ought to think about what extra features the gateway should have that one doesnt get from conventional search engines. He said that the idea as originally planned of having meta data sets could represent such added value.

Search interface options

Marianne pointed out that we ougth to make a clear distinction between what we want the output screens to look like and what we want to input form to look like, which she identified as two seperate problems. She asked whether the standard fits user needs and suggested that we need feedback from users as to whether the search form contents meets their needs.

Susanna suggested that we could have two different search options:

1) a free text search option for the general public and

2) a keywords search facility for librarians.

It was suggested that it could be useful to make an example of what a search form could look like. It was said that the search forms should support browsing as well as free text searching. In order to do free text searching, it was suggested that the search form could feature fields which have subject headers, e.g. a country field or a subject field. Marie-Lise suggested that we ought to enable users to link keywords entered using boolean logic to make searching dynamic,

e.g. Greece + women + methadone. In order to enable users to browse search terms, it was suggested that subject keywords could be presented on the search form. It was noted that one needs to have the subject headers on the inputting form first in order to be able to display them on the search form. Susanna suggested that we specifically look at how other gateways organise information.

Other examples taken from SOSIG:

The main interface of the SOSIG gateway features a subject index display similar to the top level of a classification. A click on one of the keywords (in our example: Economics") resulted in the display of a list of terms that represent subjects that are related to Economics" at the top of the page. This is followed by a listing of websites by type. The categories listed below werent discussed in detail but are given here for completeness:

Editors choice, Articles/Papers/Reports (collections), Articles/Papers/Reports (individual), Bibliographic Databases, Books/Book Equivalents, Companies, Data, Educational Materials, Government Publications, Governmental Bodies, Journals (contents and abstracts), Journals (full text), Mailing Lists/Discussion Groups, News, Organisations/Societies, Reference Materials, Research Projects/Centres, Resource Guides, Software.

Discussion of the ADIN example:

Susanna initially briefly presented some background information about this project. ADIN, the Australian Drug Information Network (ADIN) provides a central access point to Internet-based alcohol and drug information that are provided in Australia and internationally by prominent organizations. The front page provides two search options: 1) a general search which works with check boxes, thematic areas and free text search and 2) a search for professionals option. It was noted that some websites cover specifically one type of drug and that it is difficult to discriminate such a website from other websites on drugs that broadly cover drugs. Susanna said that we dont want to generate large numbers of unspecific results. An ADIN gateway search using the keyword: drug resulted in a large number of hits. This was partly due to the fact that many of the links found are actually links to various parts of the same website. A specific example here was the Lindesmith Center. There was a discussion of whether we wanted to feature hyperlinks to subsections of a website (Note: during our discussion, this was being referred to as deep linking). Susanna did not think of this as advantagous because of the large number of hits that can be generated. Also, it isnt feasible to give links to specific html pages of, or documents on one particular website because websites change all the time meaning that the updating of the gateway needs to be done more frequently than in the case of a hyperlink for an entire website. Susanna suggested that we ought to only give links to complete websites and not individual subsections of websites. Susanna proposed to compensate deeplinking by indicating names of menu sections / subsections where information on a specific topic can be found.There would be the possibility to add a relevant field in the questionnaire for this, and also to create an relevant option in the output view (search results details). This proposal needs further discussion.

Discussion of solutions to enhance better communication amongst members of the gateway project working group

Marie-Lise said that we should find a better way to communicate. She thought it insufficient to communicate by ad hoc emails. She took the view that in order to be able to discuss a particular problem we ought to work out a way of communication around specific issues. Thomas suggested that all emails should be circulated so one wont miss out on any. As a solution to this problem, Marianne suggested a virtual office which she could easily create. This would make communication easier and would allow the working group to organise times when everybody is available to discuss particular problems and issues online. There was agreement that a virtual office could also be used by editorial board of the subject gateway.

Marianne found a chat facility which is already part of the ELISAD yahoo group. The suggestion was made that we can try the Yahoo chat option. Marianne also showed us an example of an electronic forum (virtual office) she created which is located at www.annapa.com. Emails send from this package only stay in a forum and one can view messages for specific discussions. This is in contrast to Yahoo where emails are disseminated to the home computers of all participants and dont stay in a forum.

Action: Marianne will set up a virtual office using www.ANNAPA.com.

Decisions on data collection tool

Decision: Susanna will add free text fields to various thematic parts of the gateway form where evaluators can enter the location of specific thematic parts of an evaluated website if there are any. This was thought of as being an alternative to having links to various parts of the same website ( no deep linking"). The term used in the discussion to describe such thematic parts was rubric".

It was said that we need to have two search result screens:

I) One for the title + hyperlink

II) One draft screen for detailed search results which features the following fields:

1) title/publisher

2) abbreviation

3) geographic coverage

4) languages

5) subject

6) keywords

7) abstract (needs to be written, short description, information online)

8) URL

9) Field relating to subsections

User Manual:

Stephan was tasked in March 2001 to write a user manual which defines essential terms given on the data collection form and which explains in some detail how the form is supposed to be filled in. Stephan reported that he initially started writing the manual but then found it difficult to identify definitions for terms which are accepted across the board. He said that he is in no position to decide for others about the definitions to be chosen. Another problem was finding appropriate definitions in the first place. He had a look at some EMCDDA publications which include definitions but these already look dated. Stephan suggested that working group members ought to have consensus over chosen definitions. Stephan also refered to an email Susanna sent to working group members before this meeting in which she suggested that we ought to develop a pool of terminology. Stephan also reminded the working group that DrugScope is currently updating its thesaurus. He said that as part of this process, thesaurus team members identify documents that contain definitions or other useful keywords. Stephan said that he will forward any such documents to Susanna.

Susanne, Stephan and Anne will participate in writing the User Manual.

Missing website selection criteria for choosing websites

Decision: It was decided that all ELISAD members ought to collect URLs before the February meeting in Lisbon. This collection process entails making lists of websites that

a) are to be included in the gateway and

b) are not to be included in the gateway

The selection or rejection process is itself an important part of our work and needs to be documented. This will give us valuable information because the listing of included websites will enable us to estimate the work amount involved. The rejected websites listing will enable us to document the research methodology.

Thomas pointed out that the selection criteria for websites that ought to be evaluated arent clear. Thomas suggested that each participant when searching for URLs ought to indicate how he/she searches. This could show other partners which ways of finding websites are successful or not. Thomas has prepared a preliminary evaluation process document which will clarify the evaluation process. This document needs to be discussed at the next gateway meeting.

Marianne said that we should not include an evaluation if it is negative. Marie-Lise suggested that we ought to have a list as well as a selection of websites which we dont want to include and then discuss the search process when trying to find websites that ought to be evaluated.

Tasks that need to be done before the next subject gateway working group meeting in February 2002:

1) Susanna will create the gateway database

2) Susanna will update the form and do changes to the form given todays decisions

3) each participant will fill in one or two forms

4) data will have to be entred on the database

5) Susanna will create output forms

6) Susanna will create search forms

7) All data collecting participants will create a list of websites that are to be included in the gateway and a list of rejected websites that are not to be included

8) Susanne, Stephan and Anne will participate in writing the User Manual

Susanna will try to do her work well before the February meeting so that we will be able to fill in forms. The February meeting will be used to produce an overview of input forms / output forms and examples of search interfaces on the basis of the 1-2 forms that by then will have been filled in by each gateway working group member.

In the meantime, we will discuss:

? having links to various parts of the same website (deep linking")

? validation

? how can we simplify the form

? limits of selection policy and tracking themes: selection criteria not clear

? methodology of selection; websearch methodology: how do you search?

? Controlled terms (tick boxes)

? free text field: will we keep it?

? User manual: definitions of terms on forms, e.g. prevention

Re: Agenda point 2 : Detection of website offers: Introduction to Internet Detective software (Bureau Andromeda)

Marianne briefly presentation the INTERNET Detective software and distributed copies of it on a CD-ROM to working group members. Marianne explained that the INTERNET Detective software was created in the framework of the DESIRE project and makes recommendations for how to create an information gateway and to evaluate websites. This information is contained in the handbook which comes with the INTERNET Detective software.

The following steps need to be done to access the handbook:

Start CD. On the CD are two directories: IDEng and IDFra

--> one can start at: offline-o.htm in both languages.

The www.desire.org website features an online version of the handbook which is more interactive than the offline tool.