Human decisions are not a simple match of preferences or needs to actions that fulfill these necessities. Decision making is rather a complex process that is driven by a person’s preferences or needs but is also influenced by people’s personality and multiple contextual aspects.
The abundance of possibilities and the associated choices that people have to make, have
stimulated the development of software tools to assist people in this decision making process.
Recommender systems are such tools that offer people relevant suggestions for all kinds of things such as audiovisual media, web pages, offers of online shops, etc. thereby alleviating the problem of information overload on the Internet. However contextual information is often ignored or not fully exploited in the recommendation process.
This project aims to develop new algorithms that assists people in the decision making process by analyzing data of user behavior and contextual features. These algorithms go beyond existing solutions by incorporating a deep knowledge about the context of people’s information need. This knowledge will be gained by developing a framework that vouches for the gathering and analysis of large amounts of raw data and processing it into meaningful contextual information. This contextual information will be used to understand and model people’s reasoning when making decisions. The applicability of the developed algorithms will be validated by real-life scenarios with multiple test users.