Exponential Decay Function-Based Time-Aware Recommender System for e-Commerce Applications

Unlike traditional recommendation systems that rely only on the user's preferences, context-aware (CARS) consider contextual information such as (time, weather, and geographical location). These data are used to create more intelligent effective systems. Time is one of most important influential factors affect users’ preferences purchasing behavior. Thus, in this paper, time-aware investigated using two common methods (Bias Decay) incorporate time parameter with three different algorithms known Matrix Factorization, K-Nearest Neighbor (KNN), Sparse Linear Method (SLIM). The performance study based an e-commerce database includes basic user actions add cart buy. Results compared terms precision, recall, Mean Average Precision (MAP) parameters. show Decay-MF Decay-SLIM outperform Bias CAMF CA-SLIM. On other hand, Decay-KNN reduced accuracy RS context-unaware KNN.