In interactive modelling of search intent, we visualize the model made of user's interests to the user, and allow her to make changes to it using an intuitive interface. This way the user is able to direct her search through direct adjustments to the user model, instead of having to reformulate keyword queries. In this short seminar I will talk about the interactive scientific search engine SciNet, and some of the usability and modelling improvements we have introduced to it. There will be a live demo of the system as well.
Antti Kangasrääsiö is a researcher in the Probabilistic Machine Learning group in Aalto University, Finland. He conducts research on interactive user modelling in search applications.
Research group: http://research.cs.aalto.fi/pml/
Personal page: http://users.ics.aalto.fi/akangasr/