Possible supervisor: Frank van Harmelen
Smoking is one of the hardest habits to beat. Like other addictions, combating smoking requires a multi-featured solution that encourages people to change their behaviors. Although a myriad number of therapy-based approaches have been taken over the years, from one-on-one therapy, to hypnosis, to self-help literature, none have sought to change smoking habits through a computer program on a daily basis. We propose to do so using an Artificial Intelligence knowledge base system, known as uQuit, which will substitute live therapy for a computer-generated interaction.
This project joins two universities each working in its own domain, the first the knowledge engineers and the second the domain experts, with a common goal. We explain the smoking cessation protocol given by the domain experts, and on the other hand the architecture of the Knowledge base system. Having developed the system uQuit is distributed via a website, and is downloaded by more the seventeen hundred users. After three months, we have 30 users whom have indicated the uQuit has helped them quit smoking.
After the success of uQuit.nl and the development of the second version of uQuit.nl. We would like to further develop the intelligence of the application with new ambitious functions and conversion of the program to a website.
The functions that were put together in the beginning of the project:
- Natural language in interaction
- Using animation (bots)
- Making the application more intelligent based on the psychological profile
This following project could start in September 2005.