Temporal Matrix Factorization for Tracking Concept Drift in Individual User Preferences

Abstract

The matrix factorization (MF) technique has been widely adopted for solving the rating prediction problem in recommender systems. The MF technique utilizes the latent factor model to obtain static user preferences (user latent vectors) and item characteristics (item latent vectors) based on historical rating data. However, in the real world user preferences are not static but full of dynamics. Though there are several previous works that addressed this time varying issue of user preferences, it seems (to the best of our knowledge) that none of them is specifically designed for tracking concept drift in individual user preferences. Motivated by this, we develop a Temporal Matrix Factorization approach (TMF) for tracking concept drift in each individual user latent vector. There are two key innovative steps in our approach: (i) we develop a modified stochastic gradient descent method to learn an individual user latent vector at each time step, and (ii) by the Lasso regression we learn a linear model for the transition of the individual user latent vectors. We test our method on a synthetic dataset and several real datasets. In comparison with the original MF, our experimental results show that our temporal method is able to achieve lower root mean square errors (RMSE) for both the synthetic and real datasets. One interesting finding is that the performance gain in RMSE is mostly from those users who indeed have concept drift in their user latent vectors at the time of prediction. In particular, for the synthetic dataset and the Ciao dataset, there are quite a few users with that property and the performance gains for these two datasets are roughly 20% and 5%, respectively.

Recommender System Based on Temporal Models: A Systematic Review

“…In the like manner, products are constantly going in and out of popularity as time passes by. As such, the traditional RS techniques that only take into consideration the user's historical ratings and usually ignore the changes that took place over time leads to underperformance of the models and consequently produced inaccurate recommendations [12,[22][23][24][25]. To overcome the above challenges, much research has been conducted on DRSs to improve recommendation accuracy.…”

“…These include user preference, social relationships, popularity of products, seasonal changes, among others [20]. An RS that assume a static data may end up generating recommendation that does not meet the user's current need [22]. Therefore, these dynamic attributes constitute different types of concept drift in RS that need to be precisely modeled to enhance the recommendation accuracy.…”

Combined Recommendation Algorithm Based on Improved Similarity and Forgetting Curve

“…There are many references for improving the quality of recommendation by focusing on the change of interest with time, for example, Margaris and Vaz [29,30] considered the impact of the time for improving prediction quality by using datasets of movies, music, videogames and books, Chen and Lo [31,32] studied on user rating for tracking their interest shift, and Vinagre [33] used forgetting mechanisms for scalable collaborative filtering.…”

Stable Confident Rating Prediction in Collaborative Filtering

“…Here author demonstrated different matrix factorization methods like SVD, PCA and PMF. Matrix factorization [4] is important to improve the prediction quality. Author used matrix factorization technique to find change in user preferences.…”