Deep Coevolutionary Network: Embedding User and Item Features for Recommendation
This addresses the need for more accurate and flexible recommender systems by overcoming the limitations of strong parametric assumptions in existing models, though it is incremental as it builds on prior point process-based approaches.
The paper tackles the problem of capturing complex, nonlinear dynamics in user and item feature evolution for recommender systems by proposing DeepCoevolve, a deep coevolutionary network model that uses RNNs over evolving networks to define intensity functions in point processes, resulting in significant improvements in recommendation and activity prediction compared to previous state-of-the-art parametric models.
Recommender systems often use latent features to explain the behaviors of users and capture the properties of items. As users interact with different items over time, user and item features can influence each other, evolve and co-evolve over time. The compatibility of user and item's feature further influence the future interaction between users and items. Recently, point process based models have been proposed in the literature aiming to capture the temporally evolving nature of these latent features. However, these models often make strong parametric assumptions about the evolution process of the user and item latent features, which may not reflect the reality, and has limited power in expressing the complex and nonlinear dynamics underlying these processes. To address these limitations, we propose a novel deep coevolutionary network model (DeepCoevolve), for learning user and item features based on their interaction graph. DeepCoevolve use recurrent neural network (RNN) over evolving networks to define the intensity function in point processes, which allows the model to capture complex mutual influence between users and items, and the feature evolution over time. We also develop an efficient procedure for training the model parameters, and show that the learned models lead to significant improvements in recommendation and activity prediction compared to previous state-of-the-arts parametric models.