Again to clarify, the vocabulary (it's probably the wrong term to use, but it has stuck now) was designed in order to conduct a survey of material in Europeana (the survey is https://docs.google.com/forms/d/12LMyQxQTGmVN9qqxdwDGRvPWfp_iHgf_d3TzNwtBpV4/viewform) which will be distributed shortly to providers to give to their data providers. There are over 30million records in the portal, and unless we approached data providers directly we would be unable to do this type of analysis. With this in mind the aggregation team (mainly me and Francesca with help from R&D, and CERL) compiled the vocabulary on a practical level for data providers, it was not meant to be an exhaustive list of subjects and object types. We used the existing media types from EDM, with CERIF subject headings, we did try to use different vocabularies for the different media types (e.g. FIAF for VIDEO, dewey for text etc) but the links were not resolvable, which was a necessary factor to consider, in addition to this the explanations for dbpedia were in a variety of languages which made it more appealing to use say over LOC and Getty (which is not yet LOD). We tried to keep it as simple as possible with a limited hierarchy for this purpose.
I agree that we could have expanded many areas, but this would lead to confusion on the data provider side (because we had to take into account all 2200 data providers and the extreme range of topics covered), and make the survey unworkable. Coupled with this we thought we might tie in the results of the survey with the EDM collection profiles, and as the dbpedia headings are easiest to add in manually during the ingestion process this made most sense. Again not to exclude other vocabularies like UNESCO/ MICHEL/PATHS, it was purely practical reasoning... We had very successful results (including collection descriptions!) with the test pilot so we hope that will continue when the survey is sent out to all providers this week.
So a compromise had to be reached between results and vocabulary content. The vocabulary as it was created, unfortunately cannot be all things to all people. However we could work on this once the survey has been completed (in Q2 this year) and improve the range covered by the vocabulary.
Hope that gives more context!
From: Discussion list for Europeana Technical Developments [mailto:[log in to unmask]] On Behalf Of Rimvydas Laužikas
Sent: 21 January 2014 06:59
To: [log in to unmask]
Subject: Re: Europeana EDM collection profile and vocabulary for describing collections
Dear Vladimir, Dear Colleagues,
thank You for vocabulary.
I think that in creation of vocabularies very important the some things. We need decide what is the aim and the object that will described by vocabulary. If this vocabulary will used for describing of Europeana's digital objects (class - "Object type", content "text", "image", "3D", "video", "sound"), we must decide what is the next. At case of "video" it's subclases (types) of videos (advertizing film, animated film, TV film and etc.) - it's aproximatelly logical. But in case of images the next class is not types of images, but the possible content of images. It's absolutelly diffrerent class, which describe the objects from reality and can be used for classification of contents for all digital objects (e.g. archaeological site or ceramics can be in content of videos and texts too). For textual documents is used the third classification schema, where is classificated not the types of textual documents ans not the types of possible content, but the types of literaturic genres (lyrics, pamphlets, theses, essay and etc.).
Another question is using of the same criteria in the same classe.
E.g. in the classification of text we find e.g. the book and the encyclopedias (do encyclopedia is not book?), or in images, e.g.
cultural property and archaeological sites (do archaeological sites are not cultural property), or in videos, e.g TV film and documentary... and etc.
By my opinion at the good classifications we must respect this rules....
My proposal for possible schema is attached....
Vladimir Alexiev <[log in to unmask]> rašė:
>> In addition we would like to share with you parallel work done by
>> Europeana and a couple of partners on a vocabulary to describe
>> collections at a generic level:
>> The aim of this vocabulary is to cover all the objects currently
>> available in the Europeana portal.
>> Combined with the EDM collection profile this vocabulary could
>> provide a basis for future collection descriptions in Europeana.
> Hi Valentine and Robina !
> I attach an email I posted on 5 Dec 2013 at the mailing list
> [log in to unmask]
> The same is "reprinted" at the EuropeanaCreative WP2 basecamp, with a
> bit of discussion by Antoine:
> Basically, the idea is that the vocabulary for describing collections
> can be enriched easily by leveraging manually selected terms with:
> - more terms from related dbpedia categories
> - more synonyms from other LOD sources, e.g. GND and DDC