In the age of information explosion, the Recommender systems have become increasingly important and popular in supporting human decision-making. In the Recommender Systems, collaborative filtering algorithms  are one of the most popular methods to create recommendations. In the collaborative filtering algorithms, a similarity measure plays a crucial role in making the recommendation, according to which, the popular measure be used today that are all symmetrical, and therefore the default, the effect of a pair of users is the same. However, in practice, that may not be true. The more experienced a user is, the more likely he/she is to have a greater influence than the less experienced or beginner, that is, the interaction between the two users is often asymmetric. This difference can lead to bias in recommendations of recommender systems. On the other hand, the user preferences sequence can change over time, that can affecting the quality of the recommendations. In this paper, we propose a new approach, using the intensity implication measure with the user's preference factor over time in the recommender system.