Hybrid Model Rating Prediction with Linked Open Data for Recommender Systems

摘要: We detail the solution of team uniandes1 to ESWC 2014 Linked Open Data-enabled Recommender Systems Challenge Task 1 (rating prediction on a cold start situation). In these situations, there are few ratings per item and user thus collaborative filtering techniques may not be suitable. order able use content-based solution, linked-open data from DBPedia was used obtain set descriptive features for each item. compare performance (measured as RMSE) three models this cold-start situation: (using min-count sketches), (SVD++) rule-based switched hybrid models. Experimental results show that system outperforms compose it. Since taken were sparse, we clustered items in reduce dimensionality profiles.