Back to the program.
Immune-Inspired Adaptive Information Filtering
Dr Nikolaos Nanas
Centre for Research and Technology -- Thessaly
Greece
n.nanas@cereteth.gr

Abstract

Adaptive Information Filtering (AIF) seeks to provide a user with relevant information, based on a tailored representation of the user's interests, called "user profile". AIF is a challenging dynamic and multimodal optimisation problem. A user may be interested in multiple topics in parallel and a variety of changes in them inevitably occur over time. In addition to short-term variations in the level of interest in certain topics, more radical, long-term changes, such as the emergence of a new topic of interest or loss of interest in a topic, may also occur. Nootropia is a user profiling model for AIF. It is the first attempt to apply concepts from immunology to AIF. In particular it adopts Fransisco Varela's view of the immune system as a organisationally closed, non antigen driven, network of cells that reacts autonomously in order to define and preserve the organisms identity, in what is called "self-assertion". Nootropia uses a network of interacting features to represent a user's multiple interests and a process of self-organisation for adapting this network to interest changes. It is the first model to tackle multi-topic representation with a single profile structure and profile adaptation to both modest and radical interest changes. Although it has been applied so far explicitly to adaptive content-based document filtering, Nootropia is not only media independent, but potentially a general model for tackling multimodal dynamic optimisation.

Keywords

Adaptive Information Filtering, Artificial Immune Systems, Autopoiesis, Self-Organisation