Personal tools
You are here: Home → wp6  
Personlisation and Feedback
The overall goal of this work area is the development of personalisation algorithms and data structures providing quality recommendations tailored for a particular end user (consumer) or a group of end users. This goal can be subdivided in to the following goals: definition and implementation of the user model structure; definition and implementation of different feedback mechanisms and optimisation of their usage in the updating of the user model; and definition and implementation of the content selection algorithms. 

These will be based on content-based and collaborative approaches with additional capability to provide content recommendations for groups of users. The algorithms (as well as the data structures) should enable provision of online and offline content recommendations, depending on the requirements given by the user. Development of a personalised content recommender system.
  Research Focus
LIVE GUI.jpg
The main research issues can be classified into the following categories:
  • Identification of the structure of the user (consumer) model. To some extent this structure depends on the metadata standards used for descriptions of content, but also on the expected field of usage.
  • Identification of the most appropriate and efficient feedback mechanisms (explicit/implicit) and synthesis of both approaches into a common updating procedure.
  • Selection of appropriate content selection algorithms. Combination of content-based and collaborative filtering mechanisms into a more efficient hybrid recommender. Related to this issue are also the requirements for online and offline selection processes.
  • Incorporation of context- aware knowledge contained in ontologies with the personalised content recommender.

Expected Results

  • Analysis of personalisation approaches and definition of requirements for personalised content selection
  • First specification of the personalised content recommender system
  • Final specification of the personalised content recommender system
  • Report on implementation and results of the personalised recommender system