Minutes Elisad Extraordinary Gateway Group Meeting

ARCHIDO / ELISAD Project:

European Internet Gateway
on Alcohol & other Drugs

Project group meeting in Lisbon / EMCDDA,
February 6-8, 2002

Minutes ELISAD Gateway Group Meeting

Lisbon, Wednesday 6^th February 2002

Present:

Minutes:

Stephan Schulte-Nahring & Anne Singer

Agenda: Wednesday 6^th February 2002

11.00 13.00 - subcontracting procedure:

- distribution of contracts and explanations
- details on EC papers to be submitted
- working timetable and process overview

13.00 14.00 Lunch

14.00 18.00 - reports on data collection and networking process:

short reports from libraries on networking experiences and websites in countries

- Scandinavia / Sweden, Denmark, Norway, Finland: CAN /Gunnel Larsson
- UK and Paneuropean /International websites: DrugScope, Stephan
- Netherlands, Flemish Belgium: Trimbos, Toine Ketelaars
- France, Spain, Portugal, French Switzerland/Belgium: Toxibase, Marie-Lise
- Italy, Italian Switzerland, Greece: Gruppo Abele /P. Brigoni
- Germany, Austria, German Switzerland: ARCHIDO / Susanna
- Eastern Europe countries: SZU Hana Sovinova / Hungary: Susanna Prepeliczay

11.00 - Opening

- Welcome, Adelaide Duarte, EMCDDA
- Opening of meeting, Susanna Prepeliczay, Archido

11.20 - Distribution of the sub-contracts to the gateway partners

Susanna Prepeliczay distributed the sub-contracts to partners, explained some specific sub-contract items and advised on the management of expenses:

" names of new persons involved in contract work must be given to Susanna
" EC asked us to use APEX plane tickets for meetings (Saturday/Sunday night stay-over)
" ARCHIDO covers hotel costs
" taxi bills arent covered by subsistence budget
" Partners need to send the boarding passes, return tickets and receipts of plane tickets to Susanna with a request for reimbursement
" All the accounts will be managed by the accounts department of Bremen University in order to meet EC requirements
" Contracts must be signed by the responsible persons in partner organisations and returned to ARCHIDO after the Lisbon meeting. Partner organisations keep one copy.
" A partner organisations work contribution can include translations, that is why it is declared as personal costs. Translation work can also be given to external agencies.

Discussion of specific sub-contract items

" Co-financing, Art. 6 of the sub-contracts: it states that centers will make a contribution to ARCHIDO. This relates to the 30% co-financing jointly provided by all organisations involved; co-financing is the work done by a centres worker whose salary is paid by his/her employer. This is not to be declared as voluntary work!
" The declaration of work done on the project needs to be received by ARCHIDO by June 2003
" Donation / contribution in kind, Art. 6: Archido can give a receipt for a donation on request if a partner needs it for a tax declaration
" repayment, Art. 7: the contract states that if the EC ever wanted to stop the project and ask for its money to be returned, for this situation, Art. 7 was designed so that ARCHIDO will not have to pay all the money, which has been paid to project partners, back out of its own pocket. The money will have to be given back by the project participants. Archido will pay 50% of the ECs project money in 2002 and the final 50% in 2003 after having received the money from the EC.

" Interim report: each partner has to provide before July 2002 an interim report which should describe how each partner manages the identification and selection of ATOD-websites (methodology used for identifying ATOD-websites/selection process, networking process, e.g. persons contacted etc.). Susanna will compile the partners reports and produce an interim summary report for the EC. All documentation needs to be written in English. We will need to repeat this report by June 2003. Susanna will submit a template of her final report to partners before the report is finished. The template will broadly have the following structure:

1) objective
2) methods
3) process
4) result

"Thomas asked whether there is flexibility in the budget lines: Susanna answered that budget amounts cannot be changed. We are obliged to follow exactly the lines and amounts given in the budget. There is a possibility of a flexibility of 10% but this requires an official request to the EC which will give an answer after a delay of three months. Such a request would reduce other budget lines at the same time.

" Thomas mentions a problem concerning the translation of information into English. As each partner will decide on how translations are to be produced, we will get different levels of translation quality. Another concern was expressed over the very small amount of money allocated to translation costs.

" Stephan Schulte-Nähring agreed to proof-read problematic data sets

" Patrizia commented that the work amount is currently difficult to estimate because
a) partners don't know how many websites they will actually find and

b) because the number of existing websites is constantly changing. Susanna said that the preliminary estimate of websites, which allocates specific numbers of websites that need to be evaluated to partners, will be amended over time to reflect a more realistic number. If there are more websites than previously anticipated, then these additional websites can also be evaluated. However, there is no additional EC-funding for describing more than the number of websites given in the preliminary estimate.

" Susanna asks the partners to be very strict with the management of the expenses justification: not only is the EC strict but also the accounts department of Bremen University

" Stefan Quensel explained that Archido is a free association and, at the same time, part of BISDRO (Bremen Institute for Drug Research). BISDRO is a university institute, part of the University of Bremen. The EC is paying ARCHIDO, the contractor, not BISDRO who, as a participating organization, is doing the scientific monitoring.

Coffee-break. Susanna collects board passes and makes a copy of return tickets.

12.30: working timetable

Lisbon meeting is used to do major changes to the data form

Last changes to the data form should be made until the end of February 2002; we will not be able to do any major changes to the data form after the data collection process started because such changes would affect already existing descriptions which then need to be changed.

Data collection and website descriptions will start in March 2002

Interim report by partners in July 2002

Prototype of the Gateway in summer

September 2002: start of prototype evaluation. There was a discussion of

a) whether we wanted to build into the gateway prototype evaluation a gateway user consultation process and

b) at what stage this should happen.

A gateway user group would be given the task to criticize the gateway output. The number of gateway users should be fixed in September.

Search function programming takes place simultaneously with the data collection process

50% of website descriptions ought to be done before our next meeting in Bremen in September 2002

13.00 LUNCH

14.00 partners reports on data collection and networking process

1. Gunnels report:

Gunnel identified 90 Scandinavian websites that can be evaluated

Gunnel searched for Swedish, Danish and Finnish websites

Method of searching:

Gunnel used the snowball method using the links of one website to access other websites. She also found good websites by accident.

Comments:

Gunnel stated that networking will overall have advantages for CANs work.

Gunnel agreed to also identify websites in Iceland

Gunnel will more actively try to identify websites on gambling

Stakes should be contacted as well as Sirus, Norway

Outi Merilainen (Stakes) agreed to help with identifying websites. Gunnel agreed to contact her.

2. Stephan's report:

He found 151 (UK: 112, Ireland: 13, 26 Pan-European) websites in total

Method of searching:

He used the listing of UK, Irish and other European websites on drugs which is published on the DrugScope website to compile a preliminary list to which he added websites he found on Alcohol Concerns website. This listing already featured 140 websites. This number is close to the number of websites the gateway project is aiming to include.

He used the snowball method to find many additional websites.

He didn't want to start off by using any of the main search engines to avoid having to trawl through large numbers of hits. He thought that it might be worth trying out subject gateways. He used the SOSIG gateway which surprisingly doesnt include many of the key organisations in the drugs field. He also accessed authoritative websites like the EMCDDA and focal points and explored their links sections.

Comments:

Problem with inclusion criteria: Stephan didnt know whether to include websites that have only some small coverage of the misuse of drugs but which are however related, e.g. a website exclusively on hepatitis C

Pan-European and International websites:

partners should forward details about pan-European websites to Stephan

Stephan will forward his list of pan-European websites to Marianne who will put it on the ELISAD website

Anne will help collect but not analyze pan-European websites.

Stephan will more actively try to identify websites on gambling

3. Toine's report (The Netherlands):

He found 50 websites in the Netherlands and 14 websites in the Flemish-speaking part of Belgium. He got 9 more from a colleague.

Method of searching:

He used the snowball method to identify websites in the Netherlands

Toine used the All the Web search engine (http://www.alltheweb.com/) which he recommended for finding European websites. He specifically specified the country extension for Belgium in the search engine by adding be in the word filter and used the keywords addiction, drugs, alcohol and tobacco to identify websites in the Flemish-speaking part of Belgium.

Comments:

Toine wanted to know whether we include the subjects: eating disorders and sexual addiction. He commented that eating disorders have nothing to do with addiction.

He found it difficult to fill in the form

4. Thomas' report (France):

Thomas has identified 50 French websites for inclusion in the gateway.

Spain: Jose del Val of Delegacion Del Gobierno Para el Plan Nacional Sobre Drogas (DGPND) provided Thomas with 72 Spanish websites with abstracts; Thomas pointed out that DGPND could make an important contribution to the gateway.

Portugal: Thomas found 18 Portuguese websites and will seek help from the Portuguese Institute on Drugs and Drug Addiction (IPDT) to identify more

Belgium: Thomas will more actively try to identify websites in the French-speaking part of Belgium; Prospective Jeunesse will help with this.

Method of searching:

Toxibase already has a database which includes descriptions of 150 French and international websites. From this collection, 50 French websites were selected for the gateway for a preliminary listing.

Toxibase used the snowball method to identify websites in France

Comments:

Switzerland: Carla Rouge will help with identifying Swiss websites;

Problem: it often happens that one Swiss website has got three domains (French, German or Italian)

Toxibase will more actively try to identify websites on tobacco and alcohol

Further browsing and searching needs to be done using meta-engines

Questions: Thomas raised a number of questions that concern the quality of websites and the scope policy. These were:

should we collect information on websites which cover in only some detail information on AIDS/HIV or hepatitis? A problem is that many health websites only feature few sections dedicated to ATOD.

Should we include good local/regional websites or even websites that serve the needs of people in particular cities?

Should we include websites which feature databases of films that cover amongst other subjects films on drugs?

Should we include websites of pharmaceutical companies that, like in France, provide methadone?

Should we include websites of treatment centers? This could be useful because users could generate their own listings of treatment centers. A problem however is that some only describe their activities and services, but don't really provide much useful information. A discussion of this question identified two possible solution to this problem:

1) websites of treatment centers are either to be included in the gateway so that listings of treatment centers can be generated or

2) we include websites which feature directories of treatment centers, e.g. MILDT website in France

In response, Susanna pointed out that the current scope policy is only a draft and said that these problems ought to be discussed in the context of our discussion of the scope policy, after this session. So far, this discussion identified generally a need to think about what users we have in mind, who the gateway users will be and what kind of questions users would ask. In general we pointed out that the richness and substantiveness of information given are important inclusion criteria. As concerns the problem of presentations of drugs organisations which do form only a part of a large website in the general health field, we said that in many cases, it is possible to describe the AOD-related part of the website as the internet presence of this AOD-related institution and give the URL extension ("deeplink"). This procedure is only feasible for large, programmed websites, but not possible for websites with conventional frames.

5. Patrizia's report

Patrizia is in charge of collecting Italian, Greek and Swiss websites in the Italian language. She is in contact with Carla Rouge (Switzerland) and the Greek Department of Health, Athens to identify suitable websites in these countries.

She found 80 (Italy: 60, Greece: 10, Switzerland: 10) websites in total

Method of searching:

she used the Abele network as well as search engines to identify Italian websites. She made two lists: one which lists all websites and one which represents a selection of websites which she feels meet our quality criteria.

She excluded websites which have only very little information on drugs.

Comments:

she found more Italian websites than were originally anticipated

Patricia found that there are few private industry websites that are good

She would include websites of local groups if she thinks that the contents is good. Richness of information was for both, Thomas and Patrizia, the main criterion for inclusion. She pointed out that although a website may be aimed at local people, the audience however might very well be international.

Problem with Greek websites: she doesnt speak Greek so she cannot make a selection of websites or assess websites or indeed other peoples work. She doesnt know who in Greece could fill in the online form.

Patrizia will draw up a list of criteria in Italian for sub-contracted workers who need to know ELISADs requirements

contact should be established with Penelopi Vasiou (UMHRI), Greece.

she would like to have a more precise definition of what quality websites means in the scope policy. This would help her with e.g. assessing the quality of some medical websites

there are many websites on AIDS in Italy. The scope policy should help her choose the ones to be included.

Patricia would like to work with an offline version of the form. Susanna responded that we should discuss the possibility of implementing this option with our programmer Bernd. Many partners had the similar questions relating to local websites, websites on AIDS and on treatment. It will be possible to send the online form via email. One can work on it offline or save it (Netscape version 6).

6. Suzanna 's report

Germany:

she found 60 German websites

Method of searching:

she used the listings of website hyperlinks on the existing websites of ARCHIDO, BISDRO, DBDD website (Deutsche Beobachtungsstelle fur Drogen und Drogensucht, German Reitox Focal Point) and other ATOD institutions

Comments:

There are subsearch engines for social sciences which she didnt find very useful

regular search engines give loads of irrelevant search results

she found many small websites but wasnt sure whether they were worth including

Susanna found many websites and expects 80 to be selected

a colleague, ARCHIDO librarian, will help with identifying websites and data collection

Austria:

she found 32 Austrian websites

Method of searching:

she used listings of hyperlinks and the snowballing technique. Susanna also asked partners in Austria (Sabine Bruyere at Anton Proksch Institut (API)) to complement her existing pre-selection list of websites.

Comments:

there are many treatment websites

she couldnt find much on information and documentation; she found that there is a lack of ATOD information websites in Austria; she couldnt find many.

Switzerland:

she found 50 Swiss-German websites. These include web-based magazines and publishers websites. Carla Rouge (ISPA) helped her with identifying websites.

Comments:

many websites are in French and German. Many websites feature the same contents in three languages in three different domains.

Two examples:

1) The Rauchen schadet!-website (http://www.rauchenschadet.ch/)

2) The Koste-website: Susanna thought that this website presented the simplest solution to the problem where the same contents needs to be presented in different languages because it features links in 3 different languages on the same front page. These then give access to the same contents in German, French and Italian.

partners covering Switzerland need to merge their lists to avoid overlapping. This means that Susanna will pass on details about Swiss websites in Italian to Patricia and websites in French to Marie-Lise. Also, Marie-Lise and Patricia will pass on details about Swiss websites in German to Susanna.

Discussion:

Marie-Lise pointed out that a URL isnt a unique identifier of a website in cases where a website has more than one URL. She wondered whether it wasnt confusing for people to be offered many URLs for the same website. She asked whether we should have the same description three times with three different URLs. Susanna thought it easier to do it this way. She pointed out, as an example, the Koste website. We will describe its German and French version and both will be included in the gateway. However, both will have the same description in English. This description only needs to be produced once. Thats why there is a need to check who is doing which Swiss website, in order to avoid an overlapping of work. In this particular case for example, Susanna decided to analyse KOSTE, so Marie-Lise will not have to analyse the COSTE website. For all similar cases, the partners involved need to resolve such issue amongst themselves.

Another problem was that some websites dont link the same website to other domains whereas some websites link to other domains.

The website in German is: http://www.koste.ch/ (Koste Schweizerische Koordinationsstelle fur Stationare Therapieangebote im Drogenbereich).

The website in French is http://www.coste.ch/ (COSTE Centrale de coordination nationale de lOffre de Therapies residentielles pour les problemes de drogues)

7. Hana Sovinova's report

She covers Czech Republic, Poland, Slovakia, Slovenia, Romania, Bulgaria and Slovakia

She found 27 websites in total

Romania: only 1 website

Slovenia: 6 websites of which only 4 work

Bulgaria: no response yet from colleagues

Poland: she found 8 websites through search engines; no response yet from colleagues

Czech Republic: she found about 10 websites; many only have few pages, e.g. treatment centers. Some university departments publish relevant information; she found that many private individuals have drugs information on their websites but the information presented is biased, therefore not worth including.

Slovakia: 2 websites

Method of searching:

She asked colleagues in those countries whose languages she doesnt speak. She also used search engines and links on relevant websites.

Comments:

she doesnt understand some of the languages of the countries she is asked to cover

she found the search results that search engines produced dissatisfactory

Suzanna has a listing of 20 Hungarian links; she forwarded it two months ago to the Budapest Drugs Library from PRADO (Beatrix Kovacs, Agota Beniczky) to check on content and appropriateness for the gateway; as they have problems with their internet connection there is no answer yet. An alternative would be to forward it to Hanna for her colleagues to check because she doesnt speak Hungarian; some websites, which only sporadically work, will not be included in the gateway.

Thomas suggested that D. Velea (Toxibase editorial committee) could help with evaluating Romanian websites.

Discussion:

Stephan Quensel asked whether to include Serbian websites. Susanna responded that we currently shouldnt include Serbia because this country isnt listed in the EC project description. But since Susanna will go to Belgrade on a visit, she would take the opportunity to find appropriate colleagues who could help analyse Serbian websites. Patricia and Thomas agreed that if we have a contact in Belgrade, we should employ it.

A brief discussion followed about what Eastern European countries we ought to include, whether its candidate countries to EU-enlargement or countries which belong to Europe but are not in the EU. Susanna concluded that we need to be pragmatic because there is no funding available for Eastern European countries. Also, there is the language and contacts problem. We will not do work within the 18 months of the project for which we do not have the resources. However, it may be possible in the future to include other countries if we have contacts available. Anne suggested that there might be an opportunity to ask for an increase in funding for getting the countries that asked for entering the EU to join our project. This could also be a future funding option. It was suggested to ask the EMCDDA on Friday 8/2/2002 about collaboration because they have relations with Eastern European countries. It was also suggested that we check with the EC whether we can include Southern European countries, e.g. Morocco.

Conclusion of the reports session

- there are currently around 780 websites collected

- Susanna said that we have to work with feasible solutions

- listings of links dont need to be sent to Archido

17.00 Scope policy session

Marianne takes the view that we should exclude websites, even good quality websites, if they do not mention drugs and that we should focus on websites which are specifically about the misuse of drugs. She also suggested that we dont include websites which only have some small drugs coverage.

Stephan Quensel commented on the scope policy and suggested that:

- we ought to be as narrow and specific as possible because of the large number of websites we have already identified for inclusion

- we delete references to acceptable sources to make sure we include private websites. He suggested deleting the sentence under point 1.4 of the scope policy which says: "Information intended for use only by an individual or local group is unacceptable".

- we include users websites to broaden views and to avoid a situation where we would only have official views represented in the gateway. We ought to also include in the scope policy private websites, which are rich in contents and original, but be selective when actually selecting websites. Susanna responded saying that point 1.2 of the selection policy states that "websites of drug users organizations will be considered" which clearly says that private websites will be included. There was agreement that the sequence under point 1.2 is not a priority list.

- we ought to include nicotine in our inclusion criteria next to alcohol and other drugs at the beginning of the scope policy

Decision: at this stage of the project, we will only include websites which feature substantial information on the misuse of drugs and exclude everything else.

Decision: AOD will be changed to ATOD (alcohol, tobacco and other drugs)

Decision: we will add to the list of criteria: economy, trade, trafficking and marketing

Minutes ELISAD Gateway Group Meeting

Lisbon, Thursday 7^th February 2002

Present:

Minutes:

Stephan Schulte-Nahring

Agenda (amended):

Continue scope policy discussion of the previous day

- report on output drafts: search results + details view in connection to anticipated needs of the users (Marianne)

12.00 13.00

- report on database conception and properties, cataloguing functionality (Bernd)
- report and demonstration: free and field-oriented search options programming in connection with output drafts, using example datasets (Bernd / Susanna)
- general discussion on professional searching needs

13.00 14.00 Lunch

14.00 18.00 session on simplifying the input form

- reporting and comments from data collectors on experiences using the new dynamic form (partners),
- comments on the form: stakes and simplification ideas (Marie-Lise Priouret)
- drafting the User Manual (Stephen Schulte-Naehring)

16.00 Coffee break

16.30 18.00

- Discussion on the rights for each partner to access the server for up-dating.
- A technical solution has to be found to provide us this facility.

10.00 - Continuing the scope policy discussion of the previous day

The discussion mainly focused on point 3.1 and point 4 of the scope policy.

Point 3.1 of scope policy: There was a discussion as to whether commercial websites should be included in the scope policy. Stephan Quensel criticised the exclusion of commercial websites which is the current policy. Susanna said that this part of the scope policy follows the DESIRE guidelines and reminded us that the idea of the gateway was to give access to information. Anne pointed out that commercial websites could be included because they can be of interest to researchers. Stephan Quensel suggested that it should be up to the users to decide whether they want to pay for the information.

Stephan Quensel suggested to delete the first sentence of point 3.1. Commercial or fee based resources will not be considered" because it relates to the selection of websites and suggested that this could be put into the selection policy.

Point 4 of scope policy: resource description:

Susanna commented that the idea of minimum criteria for describing resources is taken from the DESIRE handbook which also includes cataloguing rules.

Decision: we want to add categories that relate to the subject area of economics

Decision: we dont want to exclude websites which cover the consequences of drug use, e.g. drugs and driving, hepatitis

Decision: Commercial websites are not to be included in the gateway except for ATOD publishers.

Susanna pointed out that we can amend the selection criteria at a later stage if we find commerical websites which we think we ought to include.

Decision: the entire point 3 is to be deleted from the scope policy and inserted into the selection criteria document. A disclaimer on our website will say that

a) we dont have a commercial interest and

b) that the views expressed on websites catalogued in the gateway do not necessarily express our opinion.

Decision: point 4 of the scope policy is to be deleted

Decision: websites published by drug users rsp. drug users organisations will be included if there is a clear statement of the initiatives mission, responsible persons and publisher evidence AND if they meet the criteria for giving substantive information pertaining to one or more subject areas (e.g. "policy", "information", "services"& .

Discussion of whether to include hyperlinks that link to subsections of complex websites (hereafter called deep-linking):

This discussion mainly centred around the issues whether or not the gateway should include links to subsections of complex websites and whether we should only describe entire websites or whether we should catalogue individual parts of a website.

Toine said that there is software which can check whether links are still working

Marianne asked whether we catalogue separate sections of websites or only entire websites. Marie-Lise takes the view that we should only describe entire websites and that we have the option to give details about certain parts of a website on the form.

Marianne asked whether we wanted AOD-websites split up into different sections. She suggested that we could index the different sections of one particular website. Marie-Lise responded saying that we should not do this because of the ensuing administrative problems of maintaining the gateway over time. She argued that there would be too much updating involved because websites change all the time. Susanna pointed out that we ought to consider the feasibility of maintaining the gateway after the project funding has come to an end. She supported Marie-Lises view arguing that we should keep the updating process feasible. Marianne, in contrast, takes the view that we ought to focus on subjects because were building a subject gateway. For user it would be useful to be directed to a specific part of a website he/she is interested in. Marie-Lise responded saying that several different sections of one website can cover the same subject area. She cited as an example the MILDT website. Information on prevention tools on the MILDT website can be found in three different places: in the new prevention tools area, the prevention booklet area and the databases area.

Toine said that if a website isnt logically structured, we nethertheless have to capture the different parts of a website. He takes the view that deep-linking would enable users of our gateway to go straight to specific pieces of information rather than just being provided with a description of what can be found on a website.

Decision: we describe for now entire websites; if we later find that this isnt enough we can change this; we should now start descriptions and later get users feedback. Susanna said that well see the limits fo the feasibility of the project once weve started putting descriptions on the database. Anne noted that we could catalogue only the ATOD section of large broadly health-related websites.

Coffee break

11.30 - Output session

Bernd Titz informed us that he checks the functionality of hyperlinks once a day over night. When links are checked, they are excluded from the browsing lists and search result. A disadvantage of this software is that it doesnt say which links aren't valid anymore. Anne proposed that we should implement a solution which could tell us which links arent working.

Mariannes presentation on the thematic ordering of our gateway

Marianne gave a presentation in which she outlined a subject approach to AOD websites. She presented an experimental draft of a gateway homepage. It is divided into frames and includes a navigation frame. The top half of the screen featured a search-by-country option. It displayed a representation of the European countries covered by the gateway giving the names of European countries arranged as a circle. Once a particular country is selected, the user is then taken to a search screen which enables him/her to browse, do free-text searches or select keywords from keyword lists for searching. The bottom half of the front page features a free-text search option, a pull-down menu from which an interactive tool can be selected, and options to search by keyword or substance, and a subject browsing search facility. The right side of the screen displays a panel which features hyperlinks that represent the different types of search options as well as information about the gateway.

Marianne suggested that we offer different types of search options: browsing subject trees, free-text searches and searches using indexing terms which can be selected from a pull-down menu

Marianne proposed to have a facility which enables users to limit searches by resources, target group and language

Marianne proposed that we use the keywords which are already on the gateway online form. These could simply be turned into keyword lists.

Marianne suggested that we have a keyword Europe to give access to pan-European websites

discussion of countries: Anne commented that she doesnt favour a countries option but would rather be able to search by language. Thomas said that we need to think about what users need. Susanna prefers having a map of Europe as a country search option. Users could click on a particular country on that map. She is against Mariannes circle because it would change its shape every time new countries are added. The question was raised whether users ought to be able to search for more than one language.

target groups: it was discussed whether we want to have target groups as a search option. Marianne favours this and suggested that the ability to limit a listing of search results by target groups enables a user to refine a search. Marianne reminded us that the current list of target groups needs to be amended.

A problem was that the field oriented search should not double the browsing lists but give users additional search options in the first line such as publisher, title, keywords, descriptions& Also, a user would not have to go through a long process to get a result.

resources search option: Marianne wondered whether this ought to include facilities like syringe exchange schemes or outpatient treatment.

Anne said that we need to decide whether we want to allow users to search by type of resource or format.

we will have to decide about the kind of structure will we use to display information on the search result form.

Susanna pointed out that we could have many more browsing lists with keywords. These could cover subjects, countries and the types of information. Each browsing list could have sub-headings with the according keywords that derive from the form. Susannas proposal for browsing lists included substances and target groups next to prevention, economy, policy, research, documentation, services, treatment. This would already mean 11 main headings (incl. countries and types of info) It was also suggested that treatment and services could be put amalgamated into one list.

Thomas suggested that we need a deadline for providing feedback on changes to the forms including keyword lists (substances and subjects), and that we cant yet discuss in detail features of the output form, the structure of the gateway/site map and the ordering of headings. Susanna, Marianne and Anne will make final decisions about the implementation of the gateway. He reminded us that we must have accurate lists of pre-coordinated indexing terms to make sure that we dont have to change them much at a later stage.

Mariannes presentation on output drafts

Marianne gave a presentation during which she compared and contrasted different output formats to show what search result screens look like in other gateways and to draw conclusions as to what it could look like in the ELISAD subject gateway. Marianne used the following subject gateways: Addiction Search, SOSIG, ADIN and Drogenguide. The search result screens showed an entry for DrugScope. She also used a draft dataset which Susanna had produced. Marianne explained what happens once a user has submitted a query. The output process after a query has been submitted to the gateway is as follows:

initially, the gateway will produce a result list. This can feature short entries with each entry giving a title, a websites hyperlink and a hyperlink to the description. The second half of the screen could feature options for limiting search results even further (resources, target groups and language). When a user clicks on the link to the description, a detailed result screen will present a selection of fields which features data generated from the ELISAD gateway database. The detailed result screen could include the following fields:

Title

Language

Type of organisation

Subject

Resources

Description summary

Comments reviewer

URL

Marianne finished her presentation with questions we have to address when deciding about the structure of the output format(s):

- Which fields do we want on the output form?

- Should the full record be displayed?

- Comments? Evaluators comments?

- What sort order do we want for displaying the listing of search results?

No decisions were taken leaving it up to Marianne, Susanna and Anne to decide on these issues.

Bernd & Susannas presentation and report on output drafts

This was mainly a presentation of the form, its fields and pull-down menus.

It wasnt possible to do searches because of server problems

The following comments on the gateway form were made:

re: languages: Hanna suggested that we could have an option other and then specify other languages in a free-text field. Hanna noticed that the languages Polish and Slovak are currently missing from the list of languages.

Marianne asked which fields are being searched when you type in cocaine. Bernd said that 5 fields are currently being searched. This is because of the structure of the form which currently features six listings with names of drugs. Marianne said that she would like to see this changed to one field only. She drew a chart which suggested how the search process could be implemented. Marie-Lise criticised that this is a form problem, not a search problem, because we have on the current version of the form substances listed in 5 different parts of the form: treatment, prevention and the thematic areas. Marie-Lise generally questioned why we have 5 lists of substances and suggested that we ought to rationalise the form so that there is only one list of substances. This prompted a discussion with participants having diverging opinions summarized as follows:

a) people take the view that we should have different thematic areas with each one having its own list of substances

b) just have one list and tick options for thematic areas

Bernds presentation of the database table

Bernd explained the technical options possible to deal with this problem indicating that, given the current structure of the form, a search on a specific drug has to search all fields that list drugs. It was after the presentation still not clear how to solve this problem.

Stephan Quensel suggested that we ought to enable free text searching and get away from having a too structured approach because it is inflexible and hard do update over time if we had to add new keywords.

Thomas suggested that we have seperate discussions around questions to do with the input and questions to do with the output. He also wondered whether the form can be adapted in the long term because it is too structured. He takes the view that we must simplify the form. He thinks that the first and third parts of the form are already well designed and identified the second part as the most problematic part of the current form which he finds too long and confusing. He is of the opinion that we should find the simplest way to deal with thematic areas. He also said that managing and updating the form are two seperate issues and that we have to find a solution for doing it.

Patrizia suggested that we have a controlled keyword thesaurus for substances and thematic keywords. She suggested that we dont use tick box options but instead have a combination of keywords and free text fields.

Marie-Lise suggested that we could have a selection of tick boxes for describing main subjects (thematic areas) and refer to a thesaurus to do indexing if necessary. She takes the view that we must simplify the thematic sections. She also said that we could, rather than using a thesaurus, use keyword lists that for keywords that are taken from the form.

Stephan Quensel responded criticising that there would be no room for a common thesaurus in the long run. He argued that inputters must be free to use new keywords. He proposed that for each subject area, we ought to offer the most precise keywords, e.g. 5 keywords, but then have an option for free text for each subject matter.

Thomas suggested that we delete information on treatment because it isnt relevant. He proposed that we review part 2 asking ourselves what the value of information is on the gateway. He suggested that tick boxes can be useful but not in all areas.

13.00 - LUNCH

14.00 : Continuing debate about simplifying the input form

Marie-Lise suggested that we simply have one list of substance keywords and one list of keywords describing thematic areas/subjects. This implies that we make a list of subject terms. A gateway user could have the option to link the keywords when searching.

Marianne proposed to use the keywords from the lists which she has already compiled

Stephan Quensel made a number of proposals:

a) he would like the system to internally map to synonyms, e.g. you type in tobacco and the system internally links to nicotine

b) output form: he would like to have a facility which gives him a list of equivalent (synonyms) expressions for a search term he has keyed in

c) basket: he would like to see a basket facility into which he can drop links he would want to explore at a later stage

d) input form: he argued that we need a short title, a line for publisher who is not identical with the organisation. We need a title for entities that dont have one. When he scrolls through search results, he wants brief information plus a field header.

Anne commented that a title field on the form could help in situations where one needs to a) distinguish the several websites of a sole producer or b) between the many producers of a single website.

Anne suggested that we need two fields, both for the foreign language name of an organisation and one field for the English name of an organisation.

Marianne said that we have free text in part III (evaluation part) which describes an organisation. The first two lines should give the key essential information. These first two lines should be presented on the output form as a search result. This would be a possible solution to this problem raised by Stephan Quensel and allow that a user can quickly skim though search results.

Refering to Mariannes comment about part III, Marie-Lise proposed to have three text fields that give summaries for:

1) publisher of a website

2) main contents of the website

3) short evaluation of the website

Susanna proposed that we could add a field for an editors comment

Marianne suggested that in order to overcome the problem where a website doesnt have a name, we could display on the search result screen two lines of the description.

Decision: we will have a website title field and a publisher field

Thomas agreed with Marie-Lises proposal to structure part II of the form as follows:

the object to the left represents a free-text field. It gives a description, e.g. you can find the following information and level of information on various topics... including factsheets and booklets etc."

Substances (keyword list)

Target groups

Subjects (keyword list giving keywords treatment, prevention, policy etc.)

Marie-Lise commented that in the above free-text field, you can point out what specific resources of information one can find on a particular website. She pointed out the problem with search engines that they you what information is available but not how to get to the information that is held in a particular part of a website.

Susanna wanted to know whether we should delete the psychactive substances tick box options of the prevention, therapy and other thematic areas of part two in order to get rid of repetitions.

Thomas said that we need to update the substance lists

Susanna asked how we will deal with synonyms

Marianne gave an example in response to Susannas question. She said that if a user seraches for nicotine, the software should map to tobacco. She gave other examples such as injecting rooms which are also called shooting galeries. Marianne made the point that the mapping mechanism implies that we have to work with a controlled vocabulary. We ought to have more subject keywords.

Stephan Quensel commented that a free-text field should be added to the subject area on the form so that were able to cover new areas for which we dont yet have keywords.

Patrizia suggested that we have a field for the name of a website

Marie-Lise criticised that the three options geographic focus of activity, country of location and the free text field for geographic focus cover more or less the same thing and ought to be rationalised.

Marie-Lise argued that we ought to delete the two categories Europe and International in the country of location pull-down menu and keep the line which gives the keywords: Regional Local National European International under type of organisation. This prompted a discussion about the meaning of what geographic focus means in the context of the web which is international.

Decision: we only keep the country of location field. The publisher summary can detail the geographic focus of activity. A template for the publisher summary field could look as follows:

Title of website

Name of publisher(s) (in native language and in English where available)

Geographic focus of activity

Decision: some of the general areas are to be taken out and be made searchable in free-text fields.

Decision: we will include two fields to include the foreign language name and the English name

Decision: we will enlarge the organisation name field

Decision: we will delete the year of foundation field in the 1^st part of the form

Decision: we only keep the categories (tick box options) that start with NGO until association / membership organisation in the Type of Organisation area on the form and instead and delete the rest. We however have to make sure that the other categories are covered in the organisation description free text field in part III of the form.

Decision: Susanna said that were not supposed to cover the thematic areas of the producers, only the website. The template of the description should include thematic areas of website.

16.00 - Coffee break

Demonstration of test data

template needs to indicate capitalisation

it will be possible to save the form as a pdf file and sent it as an email attachment; there is a consensus to have a save option for offline work

there was a question around whether changes can be made to information already submitted to the gateway; Bernd needs to write a front end for user access which is password protected.

Procedure: for the time being, a partner will need to get in contact with Susanna to ask for a change to be made and she will make any changes; Bernd will write a tool that enables partners to update records themselves in the future.

Consensus: There was a clear consensus that partners want to be able to make changes themselves.The reasons for having a system which enables partners to directly access the database were identified as being:

- partners having control over data input

- being able to do changes quickly

- freeing up Susannas time

- we have to think about the future; members will need to update records in the future so they need access to the database

SOSIG has an editors choice" icon to mark a particular quality website"; there was a discussion as to whether we would want such a facility. Marianne would like a ranking of websites using stars as a visual ranking feature. (relating to Mariannes script evaluation of websites", page 6). The six criteria each could represent one star which will show up on the output form. Marianne said that is is important to have selection criteria for the evaluation part of the form. Marianne said that you can make a link from the indexing criteria to the criteria document. We can add these to the selection criteria document and then make a free text field or tick boxes. Another option discussed was whether one couldnt simply say that a website will get a star when everything is 100% in order.

Susanna said that we could add more tick boxes to the indexing of criteria for access / user-friendlyness option.

Stephan Quensel suggested that we ought to rationalise the indexation list for types of treatment and therapy. We could cut the list down to say 15 terms and have a free text field.

Thomas suggested that we should compile lists of indexing terms for substances, subjects and also look at other databases like EDDRA.

Toine criticised the wording of please specify information on addictive behaviours". He suggested that

- all items should be in alphabetical order

- include the indexing term workaholism

- manual ought to define what treatment and prevention include

- there are no standard definitions so it will be difficult for us to define what there terms such as treatment and prevention mean

Conclusion of session:

Susanna wants to do the final changes now so that they can produce a definite form

we should develop the lists of indexing terms

Minutes ELISAD Gateway Group Meeting

Lisbon, Friday 8^th February 2002

Present:

Adelaide Duarte (EMCDDA)
Anne Singer (ELISAD)
Bernd Titz (ARCHIDO)
Gunnel Larsson (CAN)
Hana Sovinova (SZU)
Maria Cruz Cristobal, (EMCDDA)
Marianne van der Heyden (Bureau Andromeda)
Marie-Lise Priouret (TOXIBASE)
Patricia Brigoni (Gruppo Abele)
Stephan Quensel (ARCHIDO)
Stephan Schulte-Nahring (DrugScope)
Susanna Prepeliczay (ARCHIDO)
Thomas Rouault (TOXIBASE)
Toine Ketelaars (TRIMBOS)
Georges Estievenart (Director of the EMCDDA)
Jaime Berdolet (EMCDDA, general co-ordinator)
Wolfgang Goetz (EMCDDA, Reitox co-ordinator, in charge also of Eastern Europe countries)
Pedro Riberio (EMCDDA, IT co-ordinator)
Nornert Frost (EMCDDA, project manager of the epidemiological database)

Minutes:

Stephan Schulte-Nahring

Agenda:

10-18 process oriented work

10.00 - short presentation for EMCDDA staff (30min + questions) (Susanna)

11.00 12.00 promotion and evaluation

- report on gateway promotion strategy and user´s consultation concept (Anne)

12.00 13.00

- data revision: concept for an editorial board (Thomas Rouault / St. Quensel)

13.00 14.00 Lunch

14.00 15.30

- data quality enhancement / methods - Prof. Stephan Quensel, BISDRO

- reflection on guidelines: scope policy and selection criteria (Susanna)

15.30 16.00 Coffee break

16.00 18.00 training

- web searching and quality selection: internet detective training (Marianne)

Adelaide Duarte introduces the EMCDDA staff members present

Susannas outline of the ELISAD gateway project

Susanne described the history of the project, gateway concept, methodology and what users will be able to retrieve. She pointed out that a central element is networking to find out about websites and raise attention of professionals to our project and, at the same time, to involve others in it. Susanna described the work already done by networking and searches on the net. She said that there would be an evaluation of its usability once the prototype of the gateway is available in summer 2002. She pointed out that user input should allow us to finalize the gateway by the end of the summer 2002. She said that quality filters would be built in the gateway: these are experts collecting the information, an Editorial Committee, and the EMCDDA being represented by Adelaide.

Georges Estievenart: talk

G. Estievenart very much welcomed the gateway project. He pointed out that what makes this project European is the fact that it is built on the networking of European libraries. He assured us the EMCDDAs full support for the project for meetings, building, facilities and staff. He expressed his belief in the need at a European level for an Internet Gateway on drugs and that this is crucial for the EMCDDA to be a full partner. He said that there may a slight problem relating to the scope of the gateway. This concerns in particular alcohol and tobacco, which isnt part of the EMCDDAs remit. However, he doesnt think this to be a major problem because many EU member states are moving to include alcohol and tobacco in their policies and activities. The EMCDDAs annual report covers polydrug use.

G. Estievenart asked for EMCDDA staff members to be on the editorial board of our subject gateway. He specified saying that the EMCDDA does not wish to control the gateway project but he would like to see full participation and contribution for the EMCDDA.

Pedro Riberio said that he wasnt aware of this project and that he would study our documents and give feedback. He suggested that the maintenance of the project would need continuous investment over a long period. He expressed his willingness to facilitate the exchanges between EMCDDA staff and ELISAD partners concerning technical issues.

Susanna welcomed the EMCDDAs support and said that the EMCDDA participation in the project was already envisaged in the initial gateway documentation.

Anne pointed out that the EMCDDA could be very helpful with the promotion of the gateway, which, if we follow the Desire handbook, starts now. She said that the EMCDDA could for example help us to get some Gateway Users that would answer our questionnaires on their needs, on the prototype tests, and for the evaluation.

Coffee-break EMCDDA staff members leave the meeting.

Anne's description of her promotional work

Summary of main points:

Anne is in charge of the web design, gateway promotion and provides a contribution in kind. She will also work closely with Susanna and Stephan Schulte-Nahring (style guide) and will also identify websites. She will respond to user feedback. Anne pointed out that promotion starts at the beginning of the project and that we need to bring the gateway project to the attention of people who matter. She said that a user consultation process needs to be done to help us with the development and implementation of the gateway. We need to elicit users expectations and update users involved in the consultation process. Anne asked for contact details of mailing list owners to be forwarded to her so that she can send promotional information to them for distribution. Anne and Marianne will do an ELISAD presentation in Athens in March 2002 and Susanna will do an ELISAD presentation for SALIS members in April 2002.

Discussion of user involvement and gateway target group(s)

Anne will prepare questionnaires for face-to-face activities. She will follow DESIRE guidelines. She said that we need to understand who our gateway users will be. Thomas disputed that librarians are our main target group. He said that we ought to target professionals broadly and exclude the general public. Susanna responded saying that 50% of librarians as a target group is overestimated. Marianne said that we ought to target professionals including librarians. Thomas made the point that we need user feedback. We ought to start early a process of user consultation and have a spread of professionals from different backgrounds to cover our target groups. We therefore need to think about the sizes of target groups. Marianne made the point that researchers ought to be included. She takes the view that researchers and librarians are key professionals likely to use the gateway. Anne responded saying that also policy makers and public administrative personnel should be included, people who write reports for public authorities. Marianne said that we ought to differentiate between practitioners and researchers. Marie-Lise said that we ought to include students. It is Marie-Lises experience that she has many students whose enquiries are at different levels of complexity. Thomas said that the gateway also might be useful for professional training or university training. He suggested that, in order to get users for our gateway user evaluation, ELISAD board members ought to provide Anne with contacts, e.g. a university and then approach them and ask for medical students. Anne asked whether we ought to have either a focus group or a large group for our gateway user evaluation. The focus group could consist of max. 10 people and we could ask them in-depth questions. We could ask the broad group broad questions. Thomas suggested that Anne should do a listing of professionals and gateway group members will add more in order to complete the list. Marianne suggested that we could ask for a student to do a user survey as part of his/her studies. Thomas suggested that we first do the questionnaire and then decide on the size of the panel. Susanna responded saying that we first need to identify users and then use the snowballing method.

Marianne sees the advantage in having a focus group in the fact that it enables us to question users in some considerable detail. Thomas said that the snowballing method should be used for large groups and that small, carefully chosen groups ought to be used for the focus group.

Anne favours two consultations, one broad consultation and one detailed consultation. She asked gateway members to pass contacts of people who could be interested in the gateway on to her. She made is clear that these people must speak English because they need to understand the questionnaire. Marianne said that we might encounter problems when using this approach. She gave Spain as an example. In Spain, nobody fills in a questionnaire if they dont have a personal relationship to the person asking for information. Susanne said that ELISAD members must be supportive in helping users fill in the form and to mediate and promote it. She said that from experience, we would probably need to ask 1000 people in order to get a response of 100 completed questionnaires.

Marie-Lise doesnt like to just give contact details to Anne because the contact might not know Anne. She proposed that we ought to ask first people whether the allow us to pass contact details on to Anne.

Susanna said that in the context of networking between libraries, we ought to ask other librarians whether they have contacts to pass on.

Marianne said that one solution would be to use anonymous questionnaires.

Anne will send a draft of the questionnaire to ELISAD members.

Suggestion for process:

1^st identify user needs

2^nd focus on focus group for user consultation

Thomas asked us to simplify the process and agrees with Marianne.

Anne asked the gateway working group what they would like her to do and what kind of user consultation the gateway working group wants now, given the discussion above. Thomas said that although there is a need for collective decisions, people should have the authority to take the initiative and work with Susanne (project leader). He stressed that we cannot enter a process where everything needs to be discussed and agreed.

Marianne said that for the user consultation process, we ought to make a communication plan about what steps Anne is going to take so that all ELISAD members know what Anne is doing.

Marie-Lise said that we ought to try to get existing questionnaires as examples to look at and that she can give Anne the addresses of training centers which could provide us with examples of questionnaires.

Stephan Schulte-Nahrings presentation: drafting the user manual

Stephan outlined the contents and structure of the manual. His main points are based on earlier papers, submitted to the gateway working group before this meeting, in which he described his ideas for a user manual, templates for free-text fields and a style guide.

The purpose of a user manual is to present guidance on how to fill in the form giving examples, definitions of terms and cataloguing rules. A style guide could feature rules for handling particular aspects such as punctuation, spelling, acronyms, abbreviations and Americanisms. Stephan wondered whether the style guide ought to be included in the user manual or whether it should be a separate document.

Display of OHPs:

- Manual structure

- Examples of entries giving guidance on how to fill in specific form fields

Template:

Stephan had made the suggestion to use templates for free text fields. A template would consist of strings of metadata in a given order which an evaluator could use as guidance when filling in the free-text field in part III of the form.

Stephan pointed out that the advantages of a template are that it gives guidance on what to cover and when, how to structure information in a free-text field and that it would enforce consistency for all free-text field descriptions.

Definitions:

The form gives keywords such as education, prevention, website publisher which need to be defined to give evaluators an idea of what we mean by these terms.

Stephan pointed out that we ought to decide whether we have precise or broad definitions. He advocated having broad definitions because they have the advantage that they are all-inclusive. Also, people might disagree over specific definitions, so having broad definitions could save time because we wont have to argue about which definitions to include.

Discussion:

Marianne prefers the name input manual to user manual because the name conflicts with another manual we will have to produce, which will tell users what the gateway is and how to use it and which gives tips for searching.

Marianne wants users to be able to complete the form. Susanna said that we could integrate this at a later stage; initially, only ELISAD members will fill in forms and do quality control. The gateway form is not available yet for others for use and we currently do not want to make it available to others. If a completed questionnaire which describes a website is submitted by a non-ELISAD person, then that description could be included in the gateway database after a quality control check by the ELISAD person who is in charge of the particular country in which the submitted website originated.

Bernd said that having different typography is not possible

Conclusions from the discussion:

Given our decisions from the previous day, we will have three text fields in part III of the form. Each field will get a template.

in selection criteria: follow guidance in handbook and implement in template quality criteria in evaluation free text field

Separate the input manual from the gateway users' manual.

we want broad definitions

13.00 - LUNCH

14.00 Thomas: presentation on evaluation process

Thomas explained what it is, what its objectives are and outlined what should be done when records are created by evaluators. He pointed out that the two main issues for quality control are:

1) coherence of gateway selection criteria and scope policy

2) quality input on the gateway form

Main points:

having a strategy on selecting sites ensures that you not only evaluate the sites you know but find previously unknown sites

Susannas role with Bernd: ensure that if fields havent been filled in, they have to identify this and get in contact with evaluator to request further information

A procedure for updating records is needed

Thomas proposed an editorial committee. This should be a group of 3-4 persons in charge of checking records where problems can occur. He suggests that we need a simple process in order to be able to maintain the gateway project over time in the long run. He proposed that each committee member could be responsible for a specific country.

Discussion:

Marianne said that it is important to have a balanced input. An evaluation committee should

1) look at the selection of the overall number of websites

2) could do cross-search evaluations. In contrast to Thomas proposal, a committee member could check one single type of organisation for all countries. This procedure contrasts with Thomas proposal. An advantage could be a high level of quality control for one subject across different countries.

Toine thought that there may be sites which cover many subjects so more than one committee member would need to look at a single form.

Proposal: Archido should check records and pass problem records on to the committee.

Marie-Lise suggested that the committee could give answers to specific questions that evaluators encounter

access to database for updating individual records will be possible in the future on the basis of password protection.

Marianne takes the view that it is too much to go through all records and that this is why there should be a mechanism for the checking of all records first. Problem records then should be picked out for evaluation.

Thomas proposed that every participant should produce a summary in which he/she reports on the selection process, the tracking of websites and the search process. We could then discuss these summaries in a discussion forum. This could help other participants, e.g. Eastern European countries who do not know our selection criteria.

Stephan Quensel proposed that everybody could list their problems and well discuss them in Bremen in the autumn 2002. We could build up a listing of questions that users have as Frequently Asked Questions, which we could make available to others.

Susanna proposed that we could have a reference field added to the form which details how you found a particular website, what search keywords and which search engine was used.

Also, she proposed that records received by Archido could initially be check to see whether all fields have been filled in and whether there are comments in the comments field which indicate a problem. If yes, then that record could be passed on to a committee member who specializes in a particular subject

A website which doesnt extensively cover drug-related subjects however could feature interdisciplinary bibliographic databases which cover any subject to do with the misuse of drugs. For this situation, it was proposed that we index the website subject coverage only but describe the database(s).

Adelaide will ask her director about the level of support the EMCDDA intends to give us

Decision about who is going to be on the editorial committee:

One person from Toxibase, Adelaide, Marianne, Anne (observer), Hanna(limited involvement)

Not joining: Stephan Quensel, Stephan Schulte-Nahring, Gunnel

Process of identifying problem records and pass on to editorial committee and function of board:

Decisions:

problematic records are to be sent to Susanna who will collects them. We will discuss them in Bremen as a committee

well use the Annapa.com facility for discussing specific problems

we can standardise solutions to problems

16.30 - Marianne's presentations:

1^st Internet Detective

2^nd Evaluation of websites: how to search for quality on the web