FCA Modelling for Recommendation

Supervisor: Dra. A. García Serrano

The presented work is framed in the context of modelling for recommendation. Recommendation pursues offering novel and relevant content to the users.
A crucial step in this process lies in the modelling of the content to b recommended and the user preferences. The main objective of this work has been the application of Formal Concept Analysis (FCA) for this modelling step.
FCA has been successfully applied to model and organize contents. We aim to testing if the same high performance remains when it is applied to model user preferences in a recommendation scenario. Up to date, a FCA-based modelling approach has been developed and successfully tested in RepLab 2013. Once the modelling step has been covered, we have focused on its application in Recommendation. This latter step is still in development: different FCA-based recommendation algorithms have been proposed and they are being testing in different recommendation scenarios.

Angel Castellanos