--18.104.22.168 18:53, 15 November 2009 (UTC)Following our well organized meeting at Budapest and with respect to the "Athena Thesaurus" or "Athena Networked Thesaurus" there is a need to elaborate on a few problematic issues which may affect the final results of the coming test and have a risky potential to turn "success" into "failure". Specific thesaurus covers quite well specific needs which, in most cases, are segmental: Artists, Periods, materials etc. Each Thesaurus serves well the expectations of its users, i.e. is modified with respect to their needs. In terms of "Values" it means that the thesaurus includes almost all the "Values" used in its segment and its geographical location. For instance - In Italian thesaurus we may find all the "Values" of the artists that are meaningful in Italy; their works are part of a specific collection or collections etc. The value "XYZ" a well known local artist of Hungary is not necessarily "A Value" of the Italian thesaurus - Italians may have never heard of him... Keeping this fact in mind leads us to realize that we will need to enrich the subsets of the thesauri, or micro thesauri with missing "Values". I am sure that each one of us knows it and understands it for 100%. But here comes a second point - if the sources for enrichment are taken, for instance, from Hungarian thesaurus and according to this thesaurus structure "Period of Activity" is not a meta data field - it will result with an updated "Athena Networked Thesaurus" which offers only partial meta data about "XYZ". It's not a problem as long as this artist is searched and retrieved by an Hungarian user - who does not expect to look for "Period of Activity". But suppose that "XYZ" is well known in Israel also and The Israeli users expect to have "Period of Activity" as an important meta data field (and use their thesaurus respectively)- here we have a problem which will not easy to handle. Realizing such difficulties should not avoid or interfere WG4 efforts to form "An Athena Networked Thesaurus", but it must affect the method of planning the future steps: 1. There must be recommendations as to minimal and optimal meta data fields per value in each category or segment of the Networked Thesaurus". 2. There should me one language which will have all the values (English..), i.e. bilingual (English and local) should become a prerequisite. 3. The preliminary tests (phase 1)must be focused on limited number of segments or categories, each well structured. Advanced segments will focus on enrichment of values rather than implementation of new categories. 4. The WG$ management team must consider continuous maintenance and development before any "New" thesaurus is announced.
Ram Shimony, Israel, 15 Nov. 2009
--> Admin 30 November 2009
Thanks a lot for your input.
Sorry for replying with delay as we are still preparing the minutes and the list of tasks to be achieved soon. We will give you a more detailed reply and include your concerns in the minutes.
Let's keep online.