Artefact-Actor-Networks (AANs) are a theoretical model to link social networks and artefact network in order to make claims about the semantic relatedness between users and their respective artefacts. The general goal of AANs is to ease the understanding of how these to knowledge entities (Trier, 2005) are interconnected, how they influence each other and how we can make use of semantic technologies in this field of research. AANs are meant to extend social networks through a layered approach of data storage and the linkage to so called artefact networks. Artefact networks basically represent bits of data from different data sources with some semantical links between them. We hypothesize that we can use the same metrics from social network analysis in artefact networks.

In Artefact-Actor-Networks we take a closer look at the multiple levels where communication and collaboration takes place in technology enhanced working and learning settings. We break down the information about those actions, objects and users into separate layers for artefacts and users that can be analyzed separately.

The different layers can be aggregated again, so that we can deal with a consolidated or pooled network.

Semantic Relations in Artefact-Actor-Networks

In the context of Artefact-Actor-Networks there exist semantic relations between actors and artefacts (AA relation), actors and actors (ACT² relation) and between artefacts and artefacts (ART² relation). Let’s take a look at the features of semantic relations in Artefact-Actor-Networks:

ACT² relations describe the nature of relationships between involved persons. They characterise simple connections, friendships or kinships. Furthermore, they can show the kind of media people are communicating with. The Friend of a Friend (FOAF) project developed a RDF vocabulary, to express interests, connections and activities of persons. It covers constructs to describe properties of persons like firstname or surname but it is extensible to define own knows relations for each person too. Such a reference may point to virtual identities like Twitter- or a ICQ-IDs. The FOAF vocabulary is not sufficient to represent all ACT² relations in the Artefact-Actor-Network. For instance, there exists no relation to specify the kind of relation between a person to its organization. For these special semantic relationships we have to extend the vocabulary, which can easily be achieved.

ART² relations provide information on how artefacts are connected. The Dublin Core standard and the SIOC project provide currently useful expedient kinds of relations. For example artefacts can reference other artefacts with references or an artefact can be a derivative of an artefact in an earlier version with isVersionOf. Also the relations hasPart and isPartOf are useful if artefacts exist of multiple parts. Furthermore, we can reuse the relation replaces stating that an artefact replaces another one and requires saying that another artefact is required to make use of the current one (e.g. a picture within a blog entry). The SIOC project provides a set of useful relations for the context of web applications like e-mails, blogs or bulletin boards. Three important relations to reuse are replyOf, which marks an artefact as the reply of another, linksTo, expressing the linkage between two artefacts, and nextVersion, if an artefact has multiple versions and to represent the history of it.

The last type of relation we use in the concept of Artefact- Actor-Networks are AA relations. This type of relation also has many different semantic shapes. They connect actors with their respective artefacts. For each artefact there exist a set of actors which have semantic relations to it, because artefacts will always be created, modified and annotated by different actors. Therefore, AA relations give information about the kind of conection between artefacts and actors. Like for ART² relations we can use Dublin Core and SIOC to get a basic set of sematic relations. The main relations in this context are creator, publisher and rightHolder. The main relations from the SIOC Project are creator, modifier and owner. The LOM standard was created for relations between learning objects and involved entities (persons or organisations). We can use metadata from the LOM standard that describe the way of participation in a learning object. Possible shapes of values are: author, publisher, editor or initiator.

Popularity: 7% [?]