Recommendation models are vital to the success of recommender systems. The latent factor model is one of the outstanding collaborative recommendation models. It faces many difficulties due to the lack of quality and quantity of preferences observed from users. An effective approach is to use both types of user preferences, which are implicit and explicit, in the latent factor models. In this paper, we aim to propose two different latent factor models, namely SC1 and SC2, for integrating these two types of user preferences in a multi-step decision-making process, instead of just a single step as in previous studies. SC1 works well for experienced users, who have a lot of field knowledge, while SC2 is suitable for users with less domain knowledge, who do not exactly know what they want in the system.

推荐模型对于推荐系统的成功至关重要。潜在因素模型是出色的协作推荐模型之一。由于缺乏从用户那里观察到的偏好的质量和数量，它面临许多困难。一种有效的方法是在潜在因子模型中同时使用隐式和显式两种用户偏好。在本文中，我们旨在提出两个不同的潜在因素模型，即SC1和SC2，以将这两种类型的用户偏好整合到一个多步骤的决策过程中，而不是像以前的研究中那样仅一步之遥。SC1非常适合具有丰富现场知识的有经验的用户，而SC2适合具有较少领域知识的用户，这些用户不完全了解他们在系统中的需求。