The theoretical model is able to describe the chosen items replying to the questions:
Here the main problem is to explicitly declare the different roles of a single person that, in turn, can be viewed as single person, family or corporate body. Anyway, a single "person" can be referred as "the creator", "the user" or "the owner" of an item. For these reasons we think that EAC-CPF and FOAF can be used for our purposes.
Our thoughs were start with FOAF: Friend Of A Friend Ontology, we have some useful properties to describe real people (eg: Napoleone Bonaparte) and some useful general information about the person. It answer to some basic questions like: " What is the gender?" and other questions that aim to get the personal information of a person. But in our case we need informations about entities, not only real person, so we choose not to use FOAF.
EAC-CPF, from the archival context, is very useful in order to differentiate the three main types of but the most important thing is the role that a person has in relation of the item. Using EAC-CPF we can overpass this lack and add a very useful information (eg: here the questions are " What is the role of this person, releated to the item?" or " Is this person the owner of the item? If yes, what kind of owner is it?"). So, the way is to use EAC-CPF.
Go back to the questions
In our model the idea of a place is quite simple. For our purposes we need just some minor information about a place to represent it with a name (or alternative names eg: Waterloo in French, Waterlô in Wallon), a country (represented as a country code eg: IT for Italy, FR for France) and finally the couple of latitude and longitude. All of these needs can be handled using the GeoNames Ontology. The choice to use GeoNames is pushed mainly because it is more a De Facto standard.
Go back to the questions
For what concern the dates, the only question that we posed ourself is: " In which format I need to express the notion of time?". Starting from this point we posed another question i.e. " Do we need the notion of time? Or we just need the notion of date?".
Because our idea is cataloged under an historical theme, we do not care about the time, we cannot even find the time on the web (eg we cannot answer to the question "At what time in the morning the Battle of Waterloo has started?"), so our focus is just on dates. The format we need is defined by the w3Consortium under the Date and Time formats, and in particular the format DD/MM/YYYY proper of Italy.
Go back to the questions
The first question we raised is: "what we want to describe?" and we tought that the best way to do it, is to describe the concept related to the item. Of course, our model has also to describe the subject/concept proper of an item. Our mission, here, is to describe the item and the concept expressed by the item itself. We raised some questions on this matter, starting from "What is the name of the concept?", "What type of entity is this?", "What is the subject or theme of item?" or "To which category the concept belongs?" ((General belongsToCategory Army).
Some of these questions are general and fully handled with the Dublin Core Metadata Initiative and RDF(S). The DC standard is the most used because: it is the most functional, its properties are very general and it grants a very high level of abstraction. SKOS is a standard that easily define what exaclty is a concept, so we adopted it in order to give the right definition of concept. CIDOC-CRM is the best way to describe, in a general way, people linked to the concept. FOAF, instead, allows us to introduce how a concept find its material representation (following the idea of DBPedia).
Go back to the questions