ABSTRACT

Modelling multimedia content semantics for interactive learning environments: The application of COSMOS in the development of ARISTOTLE

Marios C. Angelides, Department of Information Systems and Computing, Brunel University,

The added value of multimedia arises through the use of its semantic content, in other words, according to the meaning it depicts. Modelling the semantic content of multimedia for use in applications enables both the user and the application to pose, answer and engage in interaction over questions such as "What is happening at time t? What objects are being interacted with at time t? What is the relative context of interaction (i.e. what else has, is, and will be going on)? Which media have footage of object x, event y or action z?" To achieve this, schemes like COSMOS have aimed to encapsulate a number of semantic content aspects: events, actions, temporal relationships between events and actions, objects and object properties, and spatial relationships between objects. Semantic content modelling serves to provide a representation of the content apparent within multimedia streams. Therefore, semantic content models are used as 'surrogates' for the original multimedia streams. That is to say, all queries and responses that are required for the user and system to engage in interaction are undertaken on the semantic content model, not the multimedia streams. Because of this, the semantic content model therefore has to be tightly integrated with the multimedia data streams using efficient referencing strategies. Since semantic content models serve as surrogates, the richer the representation within the model, the more useful it is to the application and thus to the user. The application of COSMOS in the development of ARISTOTLE exemplifies modelling of multimedia content semantics in interactive learning environments.
Maintained by rbennett@cs.ucl.ac.uk