ARCHIDO / ELISAD Project: European Internet Gateway on Alcohol & other Drugs Technical Meeting in Bremen

Minutes of the Bremen technical meeting, May 14-16, University Bremen

Participants:
Anne Singer (ELISAD, France), Marianne van der Heijden (Bureau Andromeda, NL), Susanna Prepeliczay (ARCHIDO, Germany), Bernd Titz (Programmer).
Virtual participant: Stephan Schulte-Naehring, DrugScope (UK)
Minutes: Susanna Prepeliczay, ARCHIDO

1^st day:

Keywords catalogue revision.

Anne Singer and Susanna Prepeliczay were starting in the morning to discuss the tables "substances", and "patterns of use" and "settings". Keywords in these sections were reviewed according to their comprehensivity, with the support from Stephan Schulte-Naehring, and the results of the Gateway users survey conducted in April. We discovered that it would be more logical to split the "substances" section into generic terms, specific terms, and substance related terms. Also, the "culture" section was edited as an own thematic section renamed "history and culture" to include related keywords. All changes and additions were simultaneously implemented to the server database (SQL tables), to the online form, and to a word copy of the form (winput form), which includes help annotations for data collection and the correct use of freetext fields and checkbox options.

Resulting from our ongoing email discussion on how to distinguish user´s searches for "treatment websites" and "treatment documents", one idea was to join these 2 related categories in the output. The according browsing category would then include website information on several therapy forms and services, next to documents (reports, articles etc) and interactive services (online counseling, online treatment)

Marianne van der Heijden joined the keywords discussion and revision after lunch. The thematic sections on references used to compare included the bilingual Toxibase Thesaurus (2001), and the old (1985) ISDD Thesaurus. The keywords in the sections "economics", "history and culture", "treatment/services", were revised. The tables "substances" and "patterns of use" were re-checked with Marianne. We decided to re-establish the keywords list for "types of organisation" in the publishers description section, as a result from the users survey. To ensure comprehensivity, several changes of wordings in the form were done. It was also evident that a more logical order of the thematic keywords sections should be made.

Virtual participant: We were supported by the live assistance from Stephan Schulte-Naehring from DugScope for proof-reading and feedback of language use, spelling, grammars and meaning in comparison to the ISDD Thesaurus, Alcohol Thesaurus, and other materials. Stephan revised our work online and gave immediate feedback vie email and phone on each thematic section.

Potential misunderstandings of terms were discovered in relation of French, German, and Dutch cultural backgrounds, and in relation with the practice and paradigms in the professional ATOD fields in these countries: e.g. the understanding of interventions like syringe exchange, or injection rooms, varied strongly to be interpreted as prevention activities, or established parts of the public care system.

2^nd day:

Technical sharing with Bernd on functions of the database

Marianne, Anne and Susanna had a fullday technical sharing and training on the database structure, functions, and search options from Bernd Titz, the programmer.

1) server database: The university server is a Unix (AIX/IBM) system like many systems applied in science and research. Therefore, the Unix related format MySQL (standardised query language) was chosen for the websites catalogue.

There is a MySQL frontend which allows to administrate the tables and the fixed data entries (keywords) locally. Each thematic category corresponds to one table (e.g. prevention_options, treatment_options, substance_options). All freetext fields are included to another table named "drogenseiten". A third table includes all options tables (keywords). The complete structure is administrated by an overall table named "zuordnung".

2) freetext fields size / offline working and saving facility:

A problem was identified concerning the size of the freetext fields: HTML does only offer 2 possibilities to display freetext fields in any questionnaire. One is the "text field". For the textfield, the size of possible input has to be specified, the maximum are 255 characters. Also, the size of lay-out of this field has to be specified. To ensure good presentation of the fields, the lay-out size of them is smaller than the input size. This leads to the not very user-friendly stake, that an evaluator will a) underestimate the size of the field in relation to his/her quantity of input (no. of characters), and also b) when writing more than 60 characters, the beginning of the sentences can not be seen on the screen. The only alternative would be to put the edit size of the text field in the same size as its prospective input (i.e. 250 characters). This would make the form look so broad in the browser, that the evaluator would have to scroll forth and hence to see it. This would not be very user-friendly also.

The 2^nd possibility for freetext entries offered by HTML is the "text area". This is a much larger field where several lines of several types can be entered. The problem with these text areas is, that they a) are not limitable, that means, any evaluator could write as many characters and rows in here as he/she wants. Another technical problem related to the properties of text areas is, that searching would take quite a long time to retrieve information entered here, from the database catalogue.

We agreed that no Gateway User would wait several minutes for a search result. That is why the freetext fields were decided to remain as they are. Only for the producer/publisher description, and for the overall website description, there will be text areas.

To compensate this technical stake, a Word form copy has been designed where the evaluator can write the texts intended to be entered in the online form freetext fields and see his/her whole entry. The online form then would only serve for the transfer of data into the database.

The "Winput form" also allows to work offline, to save a local copy of the data entered, and will feature help annotations, instructions and definitions in an integrated input manual. This encompasses also a style guidance that defines spelling and types (see 5).

3) editing / correction facility:

To allow the evaluators access to the database, a script for the edition of data named "editform.php" was created. This online facility enables the correction and/or completion of records, and to immediately insert these changes to the catalogue on the server. To chose the dataset which would have to be changed, the evaluator would use the script "selection.php" which would show a table with all records that are in the website catalogue. There, the evaluator can mark the according record and then push the button "edit". As a result he/she would obtain the record as it was entered in the online form. Corrections can be done and the data record can be sent back to the database using the "submit" button. To distinguish the editform from the finalform, the editform has a grey background. Subsequently, a password protection for the use of these tools will be created in the coming weeks.

4) search options using our database: to learn how searching functions, we did a formulation of SQL queries syntax to understand the way information is processed from the database catalogue to the result screen by a PHP script.

examples:

- a user wants to retrieve the EMCDDA website from the catalogue. She/he enters "EMCDDA" in the search field. The query syntax would be as follows:

SELECT title, publisher, description FROM drogenseiten WHERE NAME = "EMCDDA"

SQL fields SQL table name SQL data = search string

- a user wants to find information on all case management therapy in Italy. The query syntax would consist of several steps and results, as follows:

SELECT FROM drogenseiten WHERE country = "Italy" à result 1

SELECT FROM result 1 WHERE treatment_ft LIKE "case-management";

SELECT FROM treatment_options WHERE treatment = "case-management";

à result 2

SELECT FROM optionen-tabelle WHERE drogenid = id1 AND fieldoption = result 2

è final result.

The search result will then be browsed as a list of links.

We decided that another help document would be necessary for the gateway user to facilitate good searching in relation to spelling etc.

5) webpages creation: We also learned that the gateway web pages will consist of stable and dynamic parts. Stable parts will base on a HTML script defining the browsing categories, and their lay-out tags. The HTML body of each intermediate (subject related) screen will be combined with a PHP head which would dynamically display the content from the database into these categories, according to the subject, to produce a list of links /URLS.

5) style guidance: More questions were concerning the style of the data entered. To ensure consistency, all evaluators have to enter their data the same way. Anne is working on a style guidance document giving according instructions for the evaluators. These style instructions will be integarted in the word file "winputform".

Also, some characters have to be avoided on the input side in order to ensure proper functioning of the script, such as... >>, <<, , &. because some of them are also programming signs which can disturb functioning of the database commands and information processing commands.

3^rd day

website structure and ordering system / revision of keywords

We went on with the revision of keywords and their ordering in thematic sections, and according changes on the form. For the gateway website conception, Marianne had prepared a paper with a possible sitemap structure and examples of the browsing categories and their sub-sections. The relation of the ordering categories in the form to the website structure and sitemap was discussed. We found that when working on these thematic sections and sub-sections, at the same time we were working on the output structure, i.e. our keywords revision was strongly related to the browsing categories and thus the sitemap. More thematic keywords sections were revised, such as "consequences and effects of use", "patterns of use / addictive behaviours" ... and "types of information available" which was then re-named to "resources".

Overview on main results:

- summarising the "publisher and documentation" section to the types of documents and materials available on websites in (online) resources and resources providers.

- according to the GU survey: implementation of a section related to environments or settings of drug use, treatments, prevention and research.

- new ordering of the thematic sections (see below)

listing of the agreed thematic sections of the form / website browsing categories:

- psychoactive substances

- patterns of substance use & addictive behaviours

- consequences & effects

- prevention & education

- treatment & services

- policy

- economics & trafficking

- history & culture

- research

- polulations & audiences

- settings & environments

- resources & providers

Simultaneously, we set up a collection of synonyms for a search glossary to enable good retrieval of data. As there were in many cases several words expressing the same meaning, the decision was taken to add the most comprehensive word to the form keywords, and systematically list terms with synonymous contents and related contents. For example, the form keyword "tobacco" must have the search equivalents "nicotine" and "smoking" to retrieve the same data if those search strings are used. Another example is for the term "hallucinogens" in the generic substances terms, which finds only "LSD" in the specific substances terms. This concerned also the term "plant drugs" which was often refered to as "smart drugs, legal highs". Here, also specific terms were collected in the glossary. By now, for more than 40 keywords terms in the form, search synonyms were collected. This collection will go on in the further development, to be as complete as possible when the search scripting will start.

Summary of results from this technical meeting:

We have now a comprehensive keyword catalogue with logical consistency that is suitable to describe and index website contents. Doublings are avoided nearly to 100%. The wording and ordering of the form has been improved significantly. With the Word "winput"-form, there is also an alternative tool to facilitate data input. In the same time, this work on the thematic sections is a work on the ordering system of the gateway website for output, and corresponds to the sitemap of the gateway site which consists of the same categories and sub-categories to browse the information included in the catalogue.

Many notes were taken for the input instructions. Together with the work done before by Stephan Schulte-Naehring on the former user manual, and the style guidance instructions collected by Anne Singer, an input manual will be the outcome of this meeting, to be worked on through the weekend.

Data collection can finally start on Wednesday May 22. It is intended to implement all changes decided on the weekend, and then send out the materials and explanations to the project group, including: link to online form, links to edit / data correction form, Word form and input manual, and advises on the use of these materials.

50% of data (according to preliminary recherche listings) shall be collected before July.

Next step will be the creation of the stable HTML bodies of homepage and intermediate screens of the gateway site, to display the data in the catalogue on the website www.elisad.uni-bremen.de .