Behaviour-based language-interactive speaking systems

The increasing availability of large amounts of (personal) data (Big Data) and the performance boost of advanced Human Language Technologies (HLT) hold great potential for self-management and empowerment in health and wellbeing. The topic of the present proposal is the combination of text mining of Dutch written and spoken client data with the design of an intelligent, personalized Dutch spoken dialogue system (SDS) that communicates with clients in an accessible manner to facilitate their self/joint-management of health and wellbeing. The project is couched in a broad conceptualization of health and wellbeing as happiness. The goal of this proposal is to study how text mining can be used for personalizing such a system and how this can be applied to a large-scale investigation of wellbeing happiness and self-empowerment of clients.

The chosen approach is innovative in that it combines text mining, HLT and large-scale use by clients. So far text mining has mainly been based on written data, while here it is extended to Dutch spoken, complex data like interviews and dialogues, as these are increasingly being used in healthcare instead of written questionnaires. The use of spoken dialogue systems in healthcare applications is here extended to complex, personalized communication on happiness and wellbeing in the Dutch language. The proposed research will produce Dutch language technologies that are relevant for a whole range of applications in healthcare and other fields, and important insights for personalized healthcare innovation. Its potential impact thus stretches well beyond the scope of this project.



This work is part of the research programme Data2Person, which is (partly) financed by the Dutch Research Council (NWO).


  • University of Twente
  • Games for Health 
  • Readspeaker


Contact information