Exploring folksonomy and cooking procedures to boost cooking recipe recommendation

Recommender systems have gained great popularity in Internet applications in recent years, due to that they facilitate users greatly in information retrieval despite the explosive data growth. Similar to other popular domains such as the movie-, music-, and book- recommendations, cooking recipe selection is also a daily activity in which user experiences can be greatly improved by adopting appropriate recommendation strategies. Based on content-based and collaborative filtering approaches, we present in this paper a comprehensive recipe recommendation framework encompassing the modeling of the recipe cooking procedures and adoption of folksonomy to boost the recommendations. Empirical studies are conducted on a real data set to show that our method outperforms baselines in the recipe domain. © 2011 Springer-Verlag Berlin Heidelberg.