| dc.description.abstract | Ontologically Driven Adaptation is a challenging field within the area of Personalised
News. Ontologies provide a structured, semantically rich methodology for the
modelling of a domain. Personalised News is an area of Adaptive Hypermedia which
is only beginning to take shape. There is a broad range of news that can be
personalised, and a variety of ways in which a Personalised News System can be
developed. Adaptive Hypermedia can be leveraged to provide a Personalised News
Service for users who have interests in particular domains. Based on the user model,
the domain, and the system itself, the news delivered can be high level general news,
or low level detailed news, depending primarily on which type a user wishes to
receive.
This Thesis looks at the tangible differences between the results provided by two
differently modelled domain models, which represent the same information space.
One domain model is a structured view of the information space, and it contains rich
semantic meaning. By this it is meant that it has rich semantic knowledge about the
concepts residing within the domain, and relationships between those concepts. This
domain model is expressed as an ontology and will be known from here on as a Strict
Ontology. The second domain model, is a taxonomy of the information space with
little or no semantic detail of the information space. It knows nothing of the
relationships between objects other than the fact that they are related. This domain
model will from here on be known as a Loose Ontology.
This Thesis will evaluate the relative benefits of the two different approaches to
representing an information space by examining the personalised user experiences
offered. This will be carried out against the backdrop of an innovative approach to
offering personalised ontologically-driven news. | en |
