Improving RNN architectures for Contextual Venue Recommendations - Naver Labs Europe
loader image

university of glasgow logoSpeaker: Iadh Ounis professor in the School of Computing Science at the University of Glasgow and leader of the Terrier Team and Craig Macdonald, senior lecturer in Information Retrieval in the School of Computing Science at the University of Glasgow.

Abstract: Venue recommendation systems aim to effectively rank a list of interesting venues users should visit based on their historical feedback (e.g. checkins). Such systems are increasingly deployed by Location-based Social Networks (LBSNs) such as Foursquare and Yelp to enhance their usefulness to users. In this talk, we show how recent adaptations of neural networks have allowed significant progress to be achieved in venue recommendation.
In particular, we describe an adaptation of the Gated Recurrent Unit Architecture RNNs that allow to incorporate contextual information associated with the users’ sequence of checkins (e.g. time of the day, location of venues) to effectively capture the users’ contextual preferences. Further we investigate the risk & robustness of these approaches.

Date: 20th March 2019

Seminars at NAVER LABS Europe are open to the public but space is limited. Register here.

Ceci correspond à une petite biographie d'environ 200 caractéres