Hello in previous lectures we have discussed the importance of representing digital data in the proper way and, we have seen a number of tools. In this lecture, we will focus on the Europeana Data Model, that is the tool used to model the extremely rich availability of digital data regarding the European cultural heritage. Okay, so first of all what is Europeana? Europeana is an internet portal, that collects data from a number of providers, about digital objects such as books, paintings, films, all around the Europe. So, as you can see from this picture, you have a number of providers and the purpose of the Europeana Data Model, is to collect, connect and enrich the meta data, provided by the providers. To this purpose, mainly the objectives of the Europeana Data Model are to facilitate the collection, connection and the enrichment of data, coming from a number of heterogeneous providers. To provide a data model capable to transcend domain specific metadata standards yet accommodating the range of richness of community standards already adopted, by the providers, and then to facilitate the participation to the semantic Web. Just to clarify this concept let's consider an example that we will use in the remaining of the lecture. So, here we are considering the Mona Lisa, the very famous painting by Leonardo and as an example there are two providers. One is the Louvre website and the other is the Joconde website. Let's have a look to these two sites. So, this is the Louvre website and as you can see there is a picture and some information about the painting, and please notice that in this case the name of the painting is portrait de Lisa Gherardini. Now going back to the other website, the Joconde website, this is probably a richer website with three pictures, a number of information again. And look here, the title of the painting is portrait the Mona Lisa. So, you see that there could be a problem right? Two distinct names for the same object. Okay good. So in order to reach this goal, the Europeana data model, EDM, is based on a set of requirements. Here we are just listing those requirements, and then in the next slides, we are going to discuss each requirement, and the tool needed in order to implement such requirement. Okay, first requirement, distinction within the provided object. Painting, book, whatever it is and it is digital representation. Then the distinction between the object, itself, and the metadata, record describing the object. Then we want to allow for multiple records for the same object. Containing potential contradictory statements about the object. Remind that we have a number of providers, so, they can provide possibly contradictory statements. Support for object that composed by other, of other objects and finally the last three requirements. Required that European data model is based on standards. Okay? So, we will see in the next slides, we will discuss in the next slides, all those requirements. Okay, requirement one, distinction between the provided object and the digital representation. Here the idea is that, again. We have the Mona Lisa painting by Leonardo, as you can see there. And this is identified uniquely, by an identifier for the real object. That in the European data model is the provided CHO. CHO stands for Cultural Heritage Object. Now, this real object has a number of web resources, associated to it. Those digital resources, those web resources are views that describe the object. The association between the real object. And, the web resources that describe this real object, is an aggregation. An aggregation that employs the ORE, so the Object Reuse and Exchange framework. Let's have a look on, what is ORE, so as you can see here. ORE, is an Open Archive Initiative Object Reuse and Exchange and define standards for the description and exchange of aggregation on web resources. So, we have two distinct objects. The real one and the web resources associated to the real object. That are considered as a whole, through the aggregation. Observe also that in terms of standardization, this graph is an RDF that we have already seen so is complained with the resource description framework. Good. Let's see an example. So in this case, you see we have the "real" Mona Lisa painting, represented by its identifier. The provided the cultural heritage object. Then, we have an aggregation that associate to the real object. It's digital views, two web resources, and those views are the views from the perspective of a specific data provider that in this case is the Direction des musèes de France. Let's see those two views, how they look like, so this is one of the views. We can see only the hands, of Mona Lisa. And this is another of the view, in which we'll see the whole painting. Good. So, requirement number two, distinction between the object, and the metadata record describing the object. So, potentially we can say look, we all know that, this is the painting of Mona Lisa and so this is the title of the painting and we all know that Mona Lisa that the creator of Mona Lisa is Leonardo da Vinci. So, possibly we can identify, the real painting with those two data, namely Leonardo da Vinci, the creator, and Mona Lisa, the title. However, as we have see before, there could be another provider, that uses consistent information to identify the same object that are, anyway, different from the previous one. In this case, as an example, the creator is still Leonardo da Vinci, but the title of the painting is not anymore Mona Lisa but Giconde. But, both of them refer to the same object. In order to avoid this ambiguity, Europeana data model, identifies each real object, each provided cultural heritage object, with a unique identifier. And so, it supports the distinction within objects and metadata describing the object. Okay, now that we have understood, there is a real object, there are digital representations of these real object, the association between the digital representations and the real object is made by ORE aggregation. That represent the point of view of a specific provider, we can enrich the model providing other information, adding descriptive metadata. As an example we can say that, the creator, we can make explicit in the schema, that the creator of the Mona Lisa is Leonardo. And the title of this painting is portrayed the Mona Lisa dite La Giconde. Very well, in order to do that we can use the doubling core meta data initiative that we already seen in previous lecture. Let's have a look, as an example. What is the Dublin Core Metadata initiative. Okay, what is it? It is a vocabulary of 15 properties for use in resource description. And if, we go down on this page we can see that. There is the creator term and specifically the creator, the definition of the creator term is an entity primarily responsible for making the resource, indeed Leonardo was the creator of Mona Lisa. And then we have the title here. A name given to the resource, and so in this case the title is Portrait De Monalisa, remind that this is the meta data that is associated with the identifier of the real object, that is the EDM provided cultural heritage object. Good. But what are the approaches to provide descriptive metadata in the Europeana data model? There are basically two approaches. One is the object centric approach so in this case we are giving statements that directly links the object with its feature. As an example, Mona Lisa is a female, or Mona Lisa's creator is Leonardo. While in the event centric approach, the focus is on the description of the events in which the object. Has been involved. So, Mona Lisa was acquired by François 1st. Okay? Good. This, in this picture, we can see that, again, the real object Mona Lisa, The creator is Leonard the title is Portrait de Mona Lisa. There is a relationship between the painting and Francois first, but this relationship is not yet made explicit. There is a connection. Let's have a look at what EDM has met means. It means that there is a relationship between a resource and the object that have happened or have happened or happened to or have happened together with the resource under consideration. So, from this Has Met relationship, we know that there is a connection but we don't yet know that this connection is due to the acquisition of the portrait, we know that Mona Lisa is a female. We know that nowadays Mona Lisa is in Paris and that has been created between the 1503 and 1506. Look that, with respect to the previous slide, in which we only said that. Leonardo Da Vinci is the creator of the Mona Lisa painting, here we have another more general relationship, the edm has meet. So why providing two relationships, a more general one has met, there is a connect, and the more specific one is the creator, Leonardo is the creator. Because in such way you can access such data both from a general point of view so if you are interested in knowing all the agents that or events that have a relationship with Mona Lisa you simply search with the general term has met. If we want to specifically know who is the creator, in this case, you have to use the specific term, creator, as in this example. Okay. So, we can even enrich those meta data. Using what are called contextual entities. So those are metadata not directly related to the object but to other resources in the description. As an example you have agents that are persons or organizations. So you. Can figure out that as an example Leonardo is an agent. You have relationship with events that are the core class for the event centric approach then you have places, you have time spent to express time periods or dates. And then a more general skos concept that allows you to use all, the entities provided by, knowledge organization systems. Like, thesauri, classification schemes and so on and so forth, okay? So, just going back to the previous example, previously we had that the real, painting Mona Lisa. is connected to Leonardo di ser Piero da Vinci because he's the creator and because ther is a very generic relationship has met. Okay, those relationships remains, but now, instead of using a simple statement, we have an agent that represent Leonardo. And through this agent we can provide further information, like the date of birth, the date of death, the fact that Leonardo is known as Leonardo da Vinci in English while is known as Lèonard de Vinci in French language. Okay. To this purpose we use the SKOS, so you see another standard, and let's have a look to these skos:prefLabel. What does it mean? You see this is the preferred lexical label for a resource. In a given language, okay? Look, so, remind that SKOS, allows the use of knowledge organization system, the so knows KOS, such as thesauri, classification schemes, and so on and so forth. And is is a standard for the sematic Web. This slide is an example of the event centric approach that is focused on the description of events that are related to the object. We have seen in the previous slides that Francois I, the king, was related. Somehow with the Mona Lisa through the EDM has met relationship, but what is this relationship actually? Has met is a very generic relationship, right? We want to have a better understanding. So to do that actually, we introduce the event related to the acquisition of the Mona Lisa. So in this case it's clear that the agent that describe the king Francois I was present at the acquisition event of Mona Lisa, as well as the painting itself. Similarly on top of the slide we have the creation event and in this case Leonardo, that is the creator of Mona Lisa, was present at the creation event of the painting and you see that the creation event is further specified through relationships like happendAt indeed, Mona Lisa was created in Florence and occurredAt, between 1,503 and 1,506. Okay, so, concluding, the Europeana data model. Perfectly allows the coexistence of both the object-centric approach and the event-centric approach. However, the object-centric approach is more developed simply because it's based on a very well-known standards such as the Dublin Core Initiative However initiatives such, CIDOC CRM or LIDO are providing a better support also for the event-centric approach. Furthermore, we have seen that the extensive use of the RDF framework, allows the EDM model to connect to a number of, resources. On the web. In the next slide we will proceed with the analysis of the tool provided by Europeana in order to support the requirements that we have described at the beginning of this class.
