Two talks about Recommendation system
Title: Improved Recommendation of Photo-Taking Locations using Virtual Ratings - Authors: Mesut KAYA, Derek Bridge - Abstract: We consider the task of collaborative recommendation of photo-taking locations. We use datasets of geotagged pho- tos. We map their locations to a location grid using a geo- hashing algorithm, resulting in a user × location implicit feedback matrix. Our improvements relative to previous work are twofold. First, we create virtual ratings by spread- ing users’ preferences to neighbouring grid locations. This makes the assumption that users have some preference for locations close to the ones in which they take their pho- tos. These virtual ratings help overcome the discrete na- ture of the geohashing. Second, we normalize the implicit frequency-based ratings to a 1-5 scale using a method that has been found to be useful in music recommendation al- gorithms. We demonstrate the advantages of our approach with new experiments that show large increases in hit rate and related metrics. - Title: Recommending from Experience. - Authors: Francisco J Peña - Abstract: In this work we present a context-aware recommeder system that extracts contextual information from user-generated reviews. The goal from our system is to mine open-ended contextual information, i.e. non-predefined contextual information, with the goal of providing recommendations tailored to the needs of each user. We use topic modeling techniques to mine the contextual information and then we use factorization machines to produce recommendations.
Mesut KAYA and Francisco J Peña
two PhD students from Insight Centre for Data Analytics
Friday, 19 August, 2016 (All day)