This dissertation investigates novel concepts and methods for the automatic recognition of dynamic user properties using statistical supervised learning and for the integration of these properties into the dialogue management process. The aim is to model the course of the dialogue adaptive to the user. Current commercial spoken dialogue systems usually do not account for dynamic user properties...

In human-human interaction, the participants’ multimodal behaviour has impact on the interaction as a whole, and similarly in spoken human-robot interactive situations, the interlocutors’ multimodal signals seem to correlate with the user’s assessment of the interaction. We explored in more detail how some aspects of multimodal behaviour (gazing, facial expressions, body posture) can predict th...

Computers offer valuable assistance to people with physical disabilities. However designing human-computer interfaces for these users is complicated. The range of abilities is more diverse than for able-bodied users, which makes analytical modelling harder. Practical user trials are also difficult and time consuming. We are developing a simulator to help with the design and evaluation of assist...

In this paper we present a new approach to incremental user recognition in fuzzy environments where user classification is updated within an epistemological model. We extend Einhorn and Hogarth’s anchor and adjustment method ([7]), itself derived from a study of human behaviour, from the point value representation of belief and evidence to the case where belief and evidence are imprecise. We re...

In this paper, we propose a unified architecture for the creation of life-long user profiles. Our architecture combines different steps required for a user profile, including feature extraction and representation, reasoning, recommendation and presentation. We discuss various issues that arise in the context of life-long profiling.

Although there has been significant research on modeling and learning user preferences for various types of objects, there has been relatively little work on the problem of representing and learning preferences over sets of objects. We introduce a representation language, DD-PREF, that balances preferences for particular objects with preferences about the properties of the set. Specifically, we...

The goal of this research is to define a semi-automated method called Applying Patterns to Hypermedia Instructional Design (APHID) for the development of instructional hypermedia applications using established software engineering and instructional design principles. Resulting applications are evaluated using techniques common to user modelling and adaptive hypermedia. 1 Applying Patterns to Hy...

In this work we introduce event-driven grammars, a kind of graph grammars that are especially suited for visual modelling environments generated by meta-modelling. Rules in these grammars may be triggered by user actions (such as creating, editing or connecting elements) and in its turn may trigger other user-interface events. Its combination with (non-monotonic) triple graph grammars allows co...

This paper presents a multi-agent approach to design of adaptive distributed collaborative and peer help environments, which addresses a number of challenges: locating appropriate human and electronic resources depending on the help-request, motivating users to help each other, and is easily extendible "in-depth" and "in-breadth". We explore two novel approaches in user modelling: modelling of ...