Collaborative filtering (CF) recommendation approaches Sample Clauses

Collaborative filtering (CF) recommendation approaches. Collaborative filtering (CF) is a recommendation technique that bases its predictions and recommendations on the past ratings or behavior of other system users (Xxxxxxxx 2011). The assumptions behind this technique are (Bernardes 2015): first, if users agree about the quality or relevance of some items, then they will likely agree about other items; and second, users’ preferences remain stable and consistent over time. The majority of CF algorithms operate by first generating predictions of the user’s preference and then produce their recommendations by ranking candidate items by predicted preferences. Two main groups of CF algorithms exist (Felfernig 2014): memory-based algorithms (also known as kNN), and model-based algorithms (also known as latent factor models). Memory- based algorithms operate over the entire user data to make predictions (e.g., (Xxxxxx-Xxxxxxx 2009, 2010), (XxXxxxx 2005)). In contrast, model-based algorithms use the user data to build a model which is then used for recommendations (e.g., (Xxxxxx 1998), (Xxxxxxx 2004)). Memory-based algorithms are simpler, seem to work reasonably well in practice and new data can be added easily (Felfernig 2014). For this reason, the rest of this subsection focuses mainly on memory-based algorithms. Memory-based algorithms have two major branches: user-user algorithms and item-item algorithms. User-user algorithms (e.g., (Xxx 2012), (Xxxxx 2013)) use a direct algorithmic interpretation of the first assumption of CF: find other users whose past rating behavior is similar to that of the current user, and use their ratings on other items to predict what the current user will like. This type of algorithms need: